1
From Data to Decision-Making Exploring Remote Sensing Integration in Urban Planning
Practices in the St. Louis Metro Area
The June 2022 issue of the academic journal Remote Sensing featured an editorial titled
“Geospatial Understanding of Sustainable Urban Analytics Using Remote Sensing” (Sabri et al.
2022). On page one of the editorial, the authors highlight a deficit in “the current practices of
data collection, data analytics, response, and urban planning approaches in addressing
environmental challenges.” Remote sensing (RS) is a knowledge-gathering toolset that utilizes
information from sensors such as drones, satellites, and radar systems to measure characteristics
of features on the earth’s surface. RS happens to be particularly suited to understanding the
heterogeneous spatial dynamics of urban areas (Gibril et al. 2020). The Remote Sensing editorial
goes on to express how RS of the urban environment can help develop new strategies to meet the
increasingly complex challenges of a rapidly urbanizing world, beyond just environmental needs
and into new and expanding areas of applications (Sabri et al. 2022).
Urban environments are dynamic and complex places that evolve over time and generate
very heterogenous and multifaceted ecosystems. In the context of rapidly accelerating
urbanization across the globe, and with human activity causing equally accelerating
environmental change at every scale, governing urban environments, and shaping the trajectory
of their development poses challenges. These challenges are often laid at the feet of municipal
urban planners. These professionals employ systematic processes that envision, design, and
manage these urban spaces across multiple geographic and temporal scales. Planners balance the
traditional frameworks and methodologies of their practice against rapid urban expansion and a
myriad of socio-economic, environmental, and systemic factors.
2
In the context of cities growing more complex and populated each day, planners emerge as
the pivotal agents of guiding responsible urban growth and mitigating disparities present in the
community. They are tasked with fostering community resilience in the face of growing
challenges due to population, resource strain, and balancing human needs with the needs of
fragile ecosystems within and around municipalities. Their unique professional training in
balancing long-term vision with short-term interventions for urban spaces is also balanced
against the contemporary planner’s need to think in terms of urban equity, inclusivity, and
resilience (Fitzgibbons 2019; Loh and Kim 2021; Bush and Doyon 2019).
While the profession of urban planning (taking many forms) has been around as long as
humans have banded together in organized communities, the unique problems plaguing modern
communities challenge the effectiveness of traditional planning methods and highlight the need
for innovative solutions and the integration of technological advancements to enhance the
planning process. This is where RS enters the stage as a valuable tool in the modern urban
planning arsenal.
According to NASA’s Earth Data Catalogue, RS is “the acquiring of information from a
distance” (Earth Science Data Systems 2019). Popularly, it refers to data collected using sensors
that are mounted on drones, airplanes, helicopters, and satellites. These include such data
products as traditional red-blue-green (“RBG”) photography, Light Detection and Ranging
(LiDAR), and multispectral imagery. This technology enables urban planners to make well-
informed plans and identify patterns and trends in the urban environment that may not otherwise
be captured through the implementation of traditional urban planning methods. RS also enables
cost-efficient data collection over large urban areas and provides critical source data for spatial
modeling and analysis.
3
While research articles, books, and academic editorials present an ever-expanding mass of
techniques and applications for RS in urban planning, it is difficult to gauge the current level of
integration of RS technologies, techniques, and analysis within modern planning organizations
and, more specifically, within the working processes of active city planners. Research in that area
– namely, how working planners may or may not be using RS and why – is crucial to the
continued development of the RS discipline and the fulfillment of its potential within urban
planning.
One of the foundational works in this research area was that of Nancy Hoalst-Pullen and
Mark Patterson. Their article, Applications and Trends of Remote Sensing in Professional
Planning established four integration ‘barriers’ for RS in urban planning: financial, applications,
technical, and expertise barriers (Hoalst-Pullen and Patterson 2011). This thesis aims to build on
these four integration barriers by reframing them as integration ‘areas’ and positioning them in a
matrix to develop a quantitative framework for assessing the level of integration for individual
municipal organizations within the St. Louis Metropolitan Statistical Area study region (hereafter
St. Louis MSA). These four integration areas, along with overall integration, are hereafter
referred to as the F.A.T.E. Integration Matrix, and introducing this concept early is an
important context for understanding the research questions below.
Purpose
The purpose of this thesis, then, is to develop an understanding of the popular RS
applications, technologies, and techniques employed by municipal planners within the St. Louis
MSA. This thesis also aims to gauge the broader level of integration of RS within these
organizations while exploring contributing factors to this integration, such as organization
characteristics and planners’ opinions.
4
A desire to know the popular RS tools and methodologies used by urban planners in the St.
Louis MSA, along with an interest in building on Hoalst-Pullen and Patterson’s (2011) four
factors, and a more basic interest in relevant perspectives on RS that may be shared by members
of the urban planning profession, have led to the development of this study and its research
questions.
Study Significance
This study will contribute to literature in multiple ways. The first is by building on a framework
introduced by Hoalst-Pullen Patterson (2011) by taking their framework into the quantitative
analysis space and attempting to standardize a survey-based methodology for gauging RS
integration at the organization and regional levels. The second contribution to the literature is
identifying the major tools, techniques, and applications for RS in the St. Louis MSA.
Knowledge of the most prominent RS tools and strategies for planners in the study region will
help academic researchers and RS stakeholders determine the best ways to align their enablement
of RS for planners with what is currently available and preferred by these practitioners. The third
contribution to the literature is going beyond assessing the current state of RS integration to
examining potential underlying factors influencing the current RS integration levels. After
identifying a potential problem/barrier (or in this case, which integration areas may be weak or
lacking), it is important to attempt to collect data to identify significant relationships between
variables. This thesis aims to accomplish the latter by collecting demographic data about
municipalities, organization characteristics (e.g., staff members’ education and experience), and
planners’ opinions on RS to identify significant relationships between these factors and
organizations’ level of integration of RS.
5
These three contributions help fill existing gaps in the literature and can enable future
researchers to directly build on this knowledge with potential for recreating some or all these
methods to characterize and investigate RS integration within other study areas. Conducting
comparative analyses would develop an understanding of the current state of RS integration with
urban planning in multiple locations throughout the world, and challenge assumptions about
which elements of academic research (intended to enable planners) are reaching this audience of
practitioners. In summary, developing advanced applications for RS is critical to advancing the
science and practice of planning. However, it is equally important to check in with those who can
directly benefit from this innovation, as planners themselves are the ones putting these
advancements into practice. If there are barriers holding back this science from reaching
planners, it is also the responsibility of academia to identify, analyze, and reduce these
challenges.
CHAPTER 2: LITERATURE REVIEW
Literature Review Organization / Objectives
This literature review aims to achieve several objectives in understanding the integration of
RS with urban planning at the municipal level. It first provides background on planning, through
outlining the profession and providing examples of typical urban planning values, processes, and
frameworks. This review then introduces RS by defining the topic, providing examples of
popular RS data types and products, and analyzing the benefits and challenges of integrating RS
techniques and technologies within the processes of municipal planners. Finally, this literature
review explores similar literature regarding the integration of RS in urban planning, summarizes
their results, and walks through the formation of the research questions.
6
More than anything else, this literature review aims to provide a foundational understanding
of both the role of the urban planner and the ways in which RS may be used within that role. The
intent is to provide the reader with the ability to contextualize the survey results obtained from
urban planners in the St. Louis MSA and enrich their understanding of challenges, opportunities,
and pathways found in this area of study.
Urban Planning
The Role of City Planners
‘City Planner’ is a complex and amorphous title. Planners (also referred to interchangeably
as urban or municipal planners hereafter) can have a range of job descriptions depending on city
size, government type, and organization structure. A basic way of understanding the profession is
that planners are the executors of city planning. The American Planning Association Planner’s
Dictionary defines City Planning as “The decision-making process in which goals and objectives
are established, existing resources and conditions analyzed, strategies developed, and controls
enacted to achieve the goals and objectives as they relate to cities and communities” (“A
Planners Dictionary (PAS 521/522),” n.d.).
The governing authority to enact zoning regulations typically doesn’t extend to staff
planners themselves, however, both Missouri and Illinois have enacted legislation which permits
a zoning board or similar body to establish zoning districts and regulations (“Missouri Revisor of
Statutes
- Revised Statutes of Missouri, RSMo Chapter 89,” n.d.). The City of St. Louis’s Planning and
Urban Design Agency summarizes the core functions of its planners as …
“[w]ork[ing] with neighborhood residents to develop neighborhood plans;
develop modifications to the zoning ordinance and the zoning map; and make
7
recommendations to the Planning Commission on 1) blighting studies and
redevelopment plans, 2) topical plans 3) Naming/Renaming a Public Street and 4)
other major plans addressing issues important to the City of St. Louis” (“Planning
Department - About,” n.d.).
City planners are essentially that, researchers and drafters of municipal plans who aid in the
organized development of their communities. These plans can include roadway or corridor plans,
transportation plans, bike and pedestrian plans, future land use plans, comprehensive city plans,
and many others (Baer 1997; Quattro and Daniels 2022). City planners are often staff to the city
planning board, municipal planning commission, and / or the city council. Collaboration between
city staff and municipal governing bodies is key to creating plans that reflect the values and
desires of the communities they represent (Anderson 2020). Planners synthesize information
from a variety of sources when drafting plans, including relevant theories on best practices in
urban planning and land-use, comments from public input sessions and committees, and advice
from professionals such as lawyers and consultants.
The dividing line between the roles of municipal planner and zoning administrator can often
become blurred in smaller communities, especially in those municipal planning organizations
with limited staff resources. When this happens, planners can be tasked with making ground-
level decisions on the fairest and most efficient ways to enforce municipal codes, such as
architectural design guidelines for new commercial developments. In this way, planners in
smaller communities complete the zoning cycle by drafting, enacting, interpreting, and enforcing
municipal zoning codes. This involves reacting to the unique situations brought on by urban
development trends and projects in their cities (Jaffe 1984; Parolek and Crawford 2008; Talen
2012).
Most cities include zoning review and approval as part of their building permitting process.
Planners with zoning administration roles may interact with developments of every scale, from
8
approving the design and placement of a residential picket fence to reviewing civil construction
plans for a large regional airport. As municipal zoning code defines the direction and
manifestation tool for a community's development vision, planners who act as zoning
administrators can leverage their familiarity with the application of zoning code to petition city
governing bodies for strategic changes to zoning code (Wolf 2019).
Planning Tools & Processes
The effective management and implementation of sustainable development practices in
cities are the result of planners’ utilization of various tools and professional processes to help
them navigate the growing complexity of their urban environments. This section focuses on the
popular city planning tools and processes that play a role in guiding decision-making processes
and make up much of the daily responsibilities of working planners. By understanding these
typical processes and tools, a foundational understanding of the responsibilities of a planner can
be formed through which to view the benefits or limitations of the eventual integration of RS into
these workflows.
Land Use Planning - Planners attempt to determine and regulate the best and most efficient use
of land at multiple scales within the city. Zoning regulations are the medium through which land
uses are regulated in the urban atmosphere, typically through designating sets of uses permitted
on properties within zoning district types (Kim 1978). Zoning districts typically set design
guidelines to regulate the form of development on property and transitions between these
districts in the context of the entire urban landscape. These guidelines reduce conflicts between
potentially conflicting land uses, such as mitigating the negative effects of industrial land uses on
nearby schools or residential neighborhoods. The goal of land use planning is to support more
9
livable communities, foster economic development, and preserve natural and cultural resources
(Lundgren 2003).
Comprehensive Plans - A city’s comprehensive plan outlines goals and strategies that guide
planner’s decisions on land-use and design for years after the document is formally adopted
(Altschuler 2001). After adoption, stakeholders – including professional planners, elected
officials, outside consultants, and advisory committees – then lean on the comprehensive plan to
guide policy decisions at every scale (Quattro and Daniels 2022). While comprehensive plans
tend to be city-wide documents, they also lay out distinct planning areas and guiding principles
for these areas. These planning areas can include designations, such as business districts, historic
preservation areas, and transit areas.
Policy Analysis and Development - Planners constantly analyze existing policies to determine
their relevance, efficacy, and overall influence on the form of urban areas. As their dual roles
often include zoning analyst or enforcement, planners can leverage their experience applying
zoning code to development to make recommendations to municipal governing bodies on needed
changes and additions. They often draft ordinances and then herald it through the processes of
committee edits and recommendations and ultimately ratification (Mandelker 1976).
Community Engagement - Planners are tasked with integrating the needs and desires of the
community in plans and ordinance drafts. They collect those data using surveys, interviews, and
public input sessions. Public input is often offered by residents through public input time during
public hearing sessions (e.g., City Council meetings, charrette meetings). Planners can invite
members of the public to open meetings, or directly solicit input from targeted members of
specific groups or demographics. Community engagement helps planners to craft more informed
and relevant plans and ordinances and reduce administrative blind spots (Van Empel 2008).
10
Zoning & Code Compliance - Zoning regulations or “codes” influence the size, shape, and
design elements present in all kinds of development in urban space, and these regulations are
enabled by federal and state laws throughout the United States, such as those which enable
zoning ordinances for municipalities in Illinois and Missouri (“Missouri Revisor of Statutes -
Revised Statutes of
Missouri, RSMo Chapter 89,” n.d.; “65 ILCS 5/ Illinois Municipal Code.,” n.d.). These codes can
determine everything from the permitted locations of sheds or pools to the number of second-
hand retail stores which can be present within city limits. Many of these zoning regulations are
intended to reduce nuisances throughout the urban area and ease the transition between land-use
types, which may be incompatible for any number of reasons. Planners are often tasked with
drafting, managing, editing, and enforcing these zoning regulations. This task does require a
degree of administrative authority and flexibility. While zoning codes shape the city on a large
scale, they cannot account for all of the extreme variability and unpredictability of diverse urban
areas, with debates about its practical utility and potential for equitable solutions argued
extensively to this day (Serkin 2020; Krasnowiecki 1980; Gray 2022).
The examination of these five typical planning processes offers a comprehensive insight into the
framework of urban planning. These processes collectively form the backbone of daily
professional planning practice and typical responsibilities. A review of each of these processes
underscores the importance of integrating them with contemporary advancements in data
gathering, processing, and presentation. As the world advances into an era of rapid urban growth
and technological advancement, leveraging innovative data gathering and analysis will be
instrumental in fostering the kinds of planning practices that cater to the needs of diverse and
complicated communities.
11
12
Remote Sensing
Defining Remote Sensing
The United States Geological Survey (USGS) establishes a more specific definition of RS as
being “the process of detecting and monitoring the physical characteristics of an area by
measuring its reflected and emitted radiation at a distance (typically from satellite or aircraft)”
(“What Is Remote Sensing and What Is It Used for? | U.S. Geological Survey,” n.d.). The
development of a science that makes observations about distant objects through the study of their
reflected or emitted radiation has a long history for humans. All film and digital photographs can
be lumped into the category of RS products, as photographic devices measure the various
wavelengths of light reflected off the surface of objects and store those data for visualization or
analysis. Many RS sensors of the twenty-first century can capture data at wavelengths (or
“bands” of wavelengths)
outside of the visible spectrum, such as infra-red or ultraviolet radiation.
In this thesis, RS will refer to techniques and technologies that enable the observation,
measurement, visualization, and analysis of reflected radiation off objects that lie on or above the
earth’s surface, often by a device attached to an Unmanned Aerial System (UAS, a.k.a. drone),
plane, or orbiting satellite (“What Is Remote Sensing and What Is It Used for? | U.S. Geological
Survey,” n.d.; Rustamov, Hasanova, and Zeynalova 2018). RS is useful for cataloguing the
presence and location of objects on the earth’s surface, but the full power of this technology is
unlocked when used in combination with geospatial information systems (GIS), geospatial
analysis, and cartography. GIS techniques and software can analyze the absolute and relative
locations of objects on the earth’s surface and search for patterns, connections, and relationships.
13
RS is typically split into two dominant forms: active and passive. Active RS platforms
typically emit some form of radiation and capture data on the “return” to generate images or data
products. Active RS includes LiDAR which, as described further on in this section, disperses
particles from an emitter and measures characteristics of them as they return to a receiver to
generate data on the surface of objects on or above the earth’s surface. Passive RS platforms do
not emit any sort of energy to gather data. These platforms capture data from a receiver which
was emitted from some other source, such as capturing light that was generated by the sun and
has reflected off objects on the earth’s surface. This is how most aerial imagery and traditional
satellite imagery platforms operate (Jia et al. 2021; Agrawal and Khairnar 2019).
Remote Sensing Products and Technology
This study focuses on the three RS ‘platforms’ / products included in phone surveys in
HoalstPullen and Patterson (2011): Aerial Imagery/Photography, Satellite Imagery, and LiDAR –
and adds one additional platform, due to its rising popularity over the past decade: Drone
Imagery. The platforms are deliberately explained here in this order, as it reflects their respective
order of historical emergence.
Aerial Imagery/Photography (AIP) - Aerial photography has the longest history of all of the
platforms, dating back to the earliest aerial photograph of the city of Paris, France, taken from a
floating hot-air balloon in the year 1858 (“History of Aerial Photography - Professional Aerial
Photographers Association Intl,” n.d.). This technology rapidly matured during the first world
war, used for reconnaissance at many altitudes, which increased the demand for developing
better and more focused camera technology for specific aerial challenges and nuances. This trend
continued into World War II, as the technology became even more in demand (“BBC - History -
Aerial
14
Reconnaissance in World War Two Gallery” 2009).
Today, AIP is commonly employed to do everything from searching for ruins in the remote
reaches of the Amazon rainforest to mapping cities and roads. Some aerial imagery platforms
even carry the capability to capture multiple wavelengths (or “bands” of wavelengths) of light,
beyond the tradition RGB (Red-Blue-Green) color spectrum. This has opened new worlds of
analysis, including the capability to measure near infra-red light for applications, such as
vegetation and water detection (Xue and Su 2017; Xie, Sha, and Yu 2008; Bijeesh and
Narasimhamurthy 2020).
Aerial imagery boasts some of the highest spatial resolutions and pixel density per-
squareunit of all the platforms mentioned here. The spatial resolution is often sub-meter, though
many modern AIP platforms can achieve sub-foot or even centimeter-level resolution
(Benediktsson, Chanussot, and Moon 2012; Cao and Lam 1997). This can make AIP especially
valuable for the detection and identification of objects such as cars, landscaping, individual trees,
signs, and much more. AIP can be taken at a multitude of angles, including that which looks
straight down to nadir and is corrected to remove perspective and present a neutral “top-down”
view using multiple geometric corrections. Photography taken and corrected in this way is called
orthophotography (Thrower and Jensen 1976). Alternatively, imagery can be taken at oblique
angles, which is especially useful for viewing things with more of a three-dimensional
perspective, or outright generating a three-dimensional model of a space or object with enough
oblique imagery at multiple perspectives (Cheng et al. 2011; Mayer 1999). Modern aerial
imagery benefits from being largely on-demand. Organizations that can afford to charter flights
from aerial imagery providers can often choose the time of year, flight pattern, area of interest,
and angle of the imagery they are purchasing.
15
Satellite Imagery – This platform has a much more recent history than AIP. While multiple
vehicles have carried cameras to space and captured imagery, the first orbital satellite with
imagery capabilities was launched in 1959. The earth has been observed from beyond the
atmosphere for nearly sixty continuous years, with multiple families and generations of satellites
all specializing in different applications (Qian 2021; Rast and Painter 2019). This platform
captures imagery of the earth (and other celestial bodies, such as the moon) from orbital altitudes
miles above the surface. This capture altitude and methodology presents multiple key differences
from traditional aerial imagery. This imagery is captured at substantially higher above the
surface, naturally reducing the spatial resolution of produced images and presenting unique
challenges, such as frequent presence of clouds, smoke, and other atmospheric material blocking
clear views of the ground surface (Zhao, Olsen, and Chandra 2021).
Popular satellite imagery platforms include the Landsat, Sentinel, and MODIS (Moderate
Resolution Imaging Spectroradiometer) series of imaging satellites, all of which are flown and
managed in part by NASA, often in partnership with organizations such as the U.S. Geological
Survey (USGS). The United States has also led the way in making much of the data that NASA’s
fleet of satellites capture free and openly available for download on data platforms such as the
USGS EarthExplorer data portal (“EarthExplorer,” n.d.). Each of these satellites, in company
among many other satellites hosted by many countries, boast a unique combination of imaging
sensors and technology. For instance, some have sensors capable of measuring land surface heat,
while others can scan the earth’s surface for topographic shifts or are specialized to track storms
and atmospheric shifts such as hurricanes (Roy, Behera, and Srivastav 2017). While aerial
imagery is largely on-demand (though expensive), satellite imagery from platforms which orbit
around the earth are limited by their orbital frequency as to how often they can capture images of
16
the same area on the earth’s surface. The presence of atmospheric debris or cloud cover can also
make multiple ‘passes’ of satellite imagery unusable for most surface-related viewing and
analysis (Gómez-Chova et al. 2017).
Light Detection and Ranging (LiDAR) – Also known as Laser Imaging, Detection, and Ranging
– This data platform encompasses a method used to determine ranges between objects on the
earth’s surface through targeting objects with lasers and measuring the time it takes for reflected
light to return to the apparatus. Due to the scattering of reflecting laser light particles, LiDAR
creates an incredibly high-density data product referred to as a ‘point cloud,’ which resembles a
cloud of returning light data projected over three-dimensional space (Wandinger 2005).
This RS methodology was founded in the 1960s, and primarily used as a range finding
technique, before being broadened to the concept of ‘light radar,’ which saw its use expand to
detecting ranges and characteristics of reflected light from huge areas of the earth’s surface and
atmosphere (Mehendale and Neoge 2020). While LiDAR systems have a broad range of
applications, the most relevant application in this study is developing three-dimensional images
and models of both objects on the earth’s surface and the topography of the land underneath
those objects. This is typically achieved at-scale with airborne LiDAR, such as that from
apparatus attached to drones or planes. Those data types and the resulting three-dimensional
modelling capabilities are useful for a broad range of applications. These can include modelling
man-made structures, generating maps of tree canopy areas, and understand the topography of an
area at high resolutions to enable analysis such as stormwater runoff modelling (Xin Wang et al.
2020; Pittman, Costa, and Wedding 2013; Dong and Chen 2017).
Drone Imagery – The fourth and final RS platform included in this study is that of drone
imagery (DI). The term drone typically refers to unmanned aerial vehicles (UAV), which are
17
piloted remotely without any on-board human interference of passengers (Stöcker et al. 2017).
Modern drones are often battery-powered, with recreational, consumer, and military-grade
models used for a multitude of purposes. They are usually smaller than most traditional manned
aircraft, making them more maneuverable and capable of capturing imagery at both lower
altitudes and more numerous angles.
Drones may be considered the most diverse of the four RS platforms listed here, as they
can carry a wide array of imaging technologies. Drones have become very commercially
available, with many quality drones capable of imagery costing mere hundreds of dollars,
compared to the thousands (or tens of thousands) needed to charter aerial imagery / LiDAR
flights, and the millions (or billions) of dollars required to launch imagery satellites (Sheng et al.
2010). For this reason, they may be more accessible to municipalities and thus worthy of
including in this study.
Applications of Remote Sensing in Urban Planning
The following applications of RS relevant to urban planning constitute the ‘tip of the
iceberg’ of all potential and developing applications found in literature. Each of the following
application areas can contribute to an urban planner’s understanding and analysis of their subject
environment and improve their planning and decision-making.
Urban Growth & Expansion Analysis - RS data enables the monitoring and mapping of urban
growth patterns and land cover changes over time. RS provides valuable information for
assessing urban expansion, including the identification of development pressures and limits.
Multiple imagery platforms can produce imagery which can be classified or segmented – broken
out by landcover type, enabling the identification of urbanized areas through the detection and
mapping of urban landcover types and density (Jat, Garg, and Khare 2008; Ji et al. 2006).
18
Vegetation & Greenspace Analysis - RS data can assess and monitor greenspaces and vegetation
cover within urban areas. It aids in identifying green corridors, evaluating the distribution of
greenspaces, and assessing the urban ecosystem's health and connectivity. Greenspace network
analysis is an up-and-coming technique which can store and analyze highly detailed greenspace
information to provide health and vitality metrics on the entire urban green network. These
metrics can provide planners and urban decision makers with the ability to make more advanced
and informed decisions at multiple scales of greenspace (García-Pardo et al. 2022; Zhang and
Shao 2021).
Landcover Analysis - Landcover analysis leverages RS technology to identify landcover
throughout the urban area – answering the question, ‘what is actually present on the ground?’ in
these spaces. Furthermore, analyses can be performed from those data to determine the ratios
between varying landcovers (such as vegetation, pavement, bare earth, and water) and speculate
on how these ratios may affect metrics, such as quality of life and (Netzband, Stefanov, and
Redman 2007b). Many planners use these measurements to advocate for policy adjustments,
such as those that may increase greenspace or make better use of mixed development (Miller and
Small 2003; Rogan and Chen 2004).
Urban Heat Island Analysis - RS data, especially thermal imagery, helps in the analysis and
mapping of urban heat islands (UHIs). UHIs refer to the phenomenon where urban areas
experience higher temperatures compared to surrounding rural areas, primarily due to higher
concentrations of heat-trapping landcover types, such as asphalt and glass, as well as a lack of
tree canopy cover and large water bodies (Chen et al. 2006; Nichol 2005). Understanding UHIs is
essential for mitigating heat-related issues and optimizing urban design and infrastructure,
especially in a world which is almost universally recording higher temperatures and more
highheat days every year (Li et al. 2011). Understanding UHIs through examining their causes
19
and effects can support the objectives of planners to create more resilient and climate-just cities
through helping to mitigate harmful effects on vulnerable and minoritized populations (Eugenio
Pappalardo, Zanetti, and Todeschi 2023).
Disaster & Risk Assessment - Urban decision makers must face the reality that increased climatic
variability produces an associated increase in disaster risk to their inhabitants. From shortened
intervals between extreme weather events and flooding, to the heightened risk of severe air
pollution, these risks can be mitigated through informed urban planning. RS can inform planners
of the presence of landcover patterns that contribute to increased disaster risk (Im, Park, and
Takeuchi 2019). One example of RS’s value is tracking the influence of impervious areas on
urban heat island and overland water flow. Once this understanding is established, planners can
create short and long-term plans that translate into governing action that shapes the future of
impervious areas throughout the urban area (Xianwei Wang and Xie 2018).
Remote Sensing Value to Urban Planners
Integration of RS in urban planning workflows can provide a comprehensive and up-
todate view of factors affecting urban areas. This knowledge enables planners to gather valuable
spatial information on land use, land cover, and infrastructure development, which supports
evidence-based decision-making, facilitates effective urban design, and ensures optimal land use
allocation. RS can also help planners implement sustainable development strategies, preserve
natural resources, and mitigate environmental risks (Netzband, Stefanov, and Redman 2007a).
This technology can enable the monitoring and assessment of interrelated urban dynamics over
time by capturing data on urban development indicators such as urban growth patterns, land
cover changes, and environmental conditions. Many RS datasets, especially satellite and aerial
images, are found to span years or decades and can be used in the development of time series
20
analysis. Integrating a temporal component to RS analysis is especially useful in understanding
historic urban growth patterns and sprawl (Herold, Goldstein, and Clarke 2003). RS helps
planners identify areas of significant development pressure and assess the impacts of
urbanization on the natural environment (Taubenböck et al. 2012).
Furthermore, RS can enhance the analysis and understanding of urban wellness indicators
such as urban heat islands, prevalence of air pollution, and the efficacy of transportation
networks. By providing detailed and spatially explicit data, RS enables the identification of
patterns, trends, and correlations that optimize urban plans (Esch et al. 2010). Those data also
informs councils and committees on optimal policies and strategies that address urban
challenges. Furthermore, during and after the implementation of those interventions, RS can be
used to measure progress towards goals and track metrics of success such as land use conversion,
biodiversity indicators, and reduction of urban heat island (Petrou, Manakos, and Stathaki 2015;
Chen et al. 2006).
Comparable Studies
The first forays of RS into the world of urban planning and problem-solving were introduced in
the 1960s (Tatem, Goetz, and Hay 2008). In many ways, this major development in the world of
urban analysis and data gathering coincided with the American space-race, as organizations such
as NASA searched for other applications of technologies developed as huge support for their
organizations flooded in during competition with the Soviet superpower (Madry 2013).
The body of literature has overwhelmingly been directed towards the development and
review of the huge body of applications for RS in urban planning practice and toward solving
urban plights. The inception of this thesis was due to the unexplored nature of the actual adoption
21
of RS techniques and technologies for urban planning applications by planning practitioners and
within municipal planning organizations.
There is minimal literature on this topic. A 2016 article from Walter Musakwa and Adriaan
Van Niekerk entitled Earth observation for Sustainable Urban Planning: Trends and Future
Directions cite a multitude of constraints which may affect the adoption of RS (or ‘earth
observation’) for urban planning, these including financial barriers, organization constraints, or a
lack of training. The authors also conclude that while the price to access, analyze, and share RS
data may be steep, it is declining, and other non-technical barriers (social acceptance, awareness,
and organization public accessibility) are all improving over time (Musakwa and Van Niekerk
2016).
Many articles speak broadly towards the integration of geographic information systems
(GIS) topics in urban planning by practitioners. GIS can be seen as a broad umbrella which
contains the sub-discipline of RS, as RS is inextricably linked to location, and the analysis of
remotely sensed data are conducted within the GIS framework (and often on GIS software). To
that end, researchers cite the significant cost of GIS software and the development of geographic
information systems as being somewhat prohibitive to the full adoption and integration of GIS
(and consequently, RS) by the larger planning community (Drummond and French 2008; Hoalst-
Pullen and Patterson 2011).
Of the literature which directly addresses the use or integration of RS into planning processes,
Nancy Hoalst-Pullen and Mark Patterson’s 2011 article Application and Trends of Remote
Sensing in Professional Urban Planning most closely reflects the essence of the research
questions posed by this thesis and introduced the framework used in this study to measure the
integration of RS and urban planning in municipal organizations. These researchers surveyed 69
22
planning agencies throughout the Atlanta metro area, asking their representatives about their use
of RS. These planning agencies spanned multiple geographies with respondents from
municipalities, counties, and regional planning organizations. This survey asked about their use
of popular RS technologies / data products (such as LiDAR, aerial photography, and satellite
imagery). The study found that while the use of some individual technologies was prominent (i.e.
forty-two of the respondent organizations reported using aerial photography in planning), the
researchers concluded that “… the advances made by scholars conducting research using
remotely sensed data are not reaching urban planners, in spite of such scholars claiming their
work would benefit the urban planning process” (Hoalst-Pullen and Patterson 2011, 259).
Furthermore, Hoalst-Pullen and Patterson (2011) identified four categories of barriers which
blocked the critical integration of RS with planning practices: “… namely, application, technical,
expertise, and financial, that limit this transfer of knowledge [from RS researchers to planning
practitioners]” (259). The four integration barriers, recontextualized in this thesis as integration
areas, provide the basis for the framework through which this thesis will analyze the nuanced
landscape of RS integration in urban planning throughout the St. Louis MSA. To paraphrase,
these four integration areas are:
1. Financial: The organization's capacity to purchase and/or maintain access to RS
imagery, data products, education, and/or software, as well as their ability to hire and
sustain employees with sufficient experience to take advantage of these resources.
2. Applications: The degree to which applications of RS technologies, techniques, and
analysis assist in fulfilling the organization’s need to complete a diverse set of duties
efficiently and effectively.
23
3. Technical: The ease with which the organization can/has integrate(d) RS into their
operations, through existing frameworks for spatial data such as GIS systems and ability
to process RS data.
4. Expertise: The level of formal knowledge, training, and education resources within the
organization for use of RS techniques and access to RS technologies and databases.
In the applications category, Hoalst-Pullen and Patterson (2011) found that while there was
widespread use of multiple RS products, aerial imagery was the only remotely sensed data
product used by cities, mainly for planning and map updating. In the technical integration
category, the authors reported that responding agencies strongly prefer vector-type data (points,
lines, polygons) over the use and manipulation of raster type data (pixel-based), which most RS
data products consist of, though remotely sensed data was reported as increasing in value to
planning agencies over the years (Hoalst-Pullen and Patterson 2011, 256). Regarding expertise,
the authors reported that formal training in RS led to increased use of satellite imagery and
LiDAR, with the organizations reporting no use of RS indicating lack of experience as their
biggest barrier (HoalstPullen and Patterson 2011, 256). Financial barriers seemed to be less of a
factor for municipal organizations, who often had access to cheap or free imagery from other
sources (Hoalst-Pullen and Patterson 2011). Though eight of the respondent agencies (undefined
agency type) still cited financial constraints as their chief constraint against using RS, the authors
concluded that it was the least constraining dimension out of the four integration barriers (Hoalst-
Pullen and Patterson 2011, 259).
One statement stands out from the Hoalst-Pullen and Patterson (2011) article, and it is at the
heart of the motivation for this study, specifically, the gap between the booming academic
literature on the applications of RS for urban planning and the actual adoption of this technology
by planning practitioners:
24
“Until the adoption of remotely sensed data acquisition and use is widespread,
technical expertise in RS is commonplace, and applications of remotely sensed
data are integrated into urban planning and sustained development decisions, the
gap between urban planning within local governments and urban planning in
academia will widen” (258).
Literature Review Conclusions
Findings in the literature review emphasize the benefit of integrating RS and urban
planning in enhancing the efficiency, accuracy, and sustainability of urban planning processes.
RS demonstrates a diverse set of applications in planning, including urban growth monitoring,
urban heat island analysis, air quality monitoring, transportation planning, and green space
assessment, among many others (García-Pardo et al. 2022; Netzband, Stefanov, and Redman
2007a; Li et al. 2011). The integration of RS in urban planning is crucial in continuing to develop
modern strategies towards optimizing urban development for human and well-being, while
preserving the ability of natural areas to thrive within and without the urban periphery (Esch et
al. 2010; Xianwei Wang and Xie 2018). The literature review also highlights the benefits of
integrating RS in urban planning, including improved data accuracy and spatial coverage, cost-
effectiveness, timely data acquisition, and enhanced visualization and analysis capabilities.
However, challenges such as data availability and access, image interpretation accuracy,
technical skills, and integration with existing planning frameworks need to be addressed to fully
harness the potential of RS technology in urban planning (Benediktsson, Chanussot, and Moon
2012; Yin et al. 2021; Masser 2001).
25
The integration of RS technology in urban planning is of utmost importance for
developing cutting-edge planning workflows for understanding the heterogenous urban landscape
and creating more sustainable, resilient, and livable cities (Im, Park, and Takeuchi 2019; Yang
2021). More specifically, it empowers planners with valuable information for evidence-based
decision-making, enhances understanding of urban dynamics, and supports the development of
effective strategies for land use allocation, environmental management, and transportation
planning (Almeida et al. 2005; Rahman 2007; Lakschmana Rao 1996). As technology continues
to advance, further research and collaboration are needed to overcome the challenges and unlock
the full potential of RS in urban planning practice.
While most research on RS for urban areas concludes that it is inherently valuable to developing
an understanding of complex urban areas and dynamics, there is a shortfall in some research
areas. Exploration of the applications of RS for urban information gathering and decisionmaking
is plentiful, but there is significantly less research on the actual adoption (or lack thereof) of RS
practices by planning professionals, especially at the municipal planning level. Understanding
the current state of this integration and identifying barriers/challenges to increasing this
integration is relevant to the continued innovation of modern planning practice, and the direct
purpose of this study (Zhu et al. 2019; “Remote Sensing and Urban Planning – A Common
Future? | Sensors and Systems,” n.d.).
RS is an incredibly powerful tool for understanding urban morphology and development
indicators, spanning the built-up, natural, and social environments. This technology provides an
unparalleled level of information about the urban environment, often accessible from the comfort
of a planner’s home or office. In many aspects, it is the future core of urban planning
informationgathering, especially in the face of environmental issues such as climate change,
26
which knows no limit of geographic scale. By understanding barriers to further integration of RS
in urban planning organizations, we can begin to remove them. Removal of these barriers will
lead to an increased organization capacity to meet challenges through a deeper understanding of
the complex and heterogeneous urban landscape.
Research Questions
1. What are the popular RS techniques, technologies, and applications used across the St. Louis
Metropolitan Statistical Area (MSA) for urban planning?
2. Do planners in municipal organizations across the St. Louis MSA believe that RS is valuable
to the urban planning process?
3. What is the current level of RS and planning integration across the St. Louis MSA, based on
the F.A.T.E. Integration Matrix?
4. Do demographic qualities of a municipality or characteristics of the municipal planning
organization have a measurable effect on an organization’s level of RS integration in urban
planning?
5. Do either organization characteristics or integration scores have a significant relationship
with organization opinions on the integration of RS and urban planning?
CHAPTER 3: METHODS
Overview
This section describes the approach used to investigate the integration of RS with urban
planning processes at municipalities across the St. Louis MSA. This study employed a
surveybased research design to gather comprehensive data pertaining to organization
27
characteristics, popular RS tools, and their utilization in planning, integration of RS, and
planning at the organization level, and planners’ opinions toward RS within the planning process.
Study Area & Subjects
Figure 1: St. Louis MSA Overview Map with County Names
The study area for this thesis was the Metropolitan Statistical Area of St. Louis (Figure 1).
The U.S. Census Bureau defines a Metropolitan Statistical Area (MSA) as consisting of at least
one urbanized ‘core area’ with over 50,000 residents and outlying counties or urban centers with
a minimum population of 10,000 people who rely on commuting to or from the ‘core area’ (US
Census Bureau “About” 2021). According to the 2021 TIGER line census shapefiles, the St.
Louis MSA contains 309 recognized “places,” which include towns, cities, and village areas (US
Census Bureau “Mapping Files” 2021). According to 2021 census estimates, the St. Louis MSA
contains around 2,810,000 residents (US Census Bureau, n.d.-a). The area covers 8082 sq. miles,
28
split evenly between the two states (IL 52.3%, MO 47.7%). The St. Louis MSA, as defined by
the
Census, consists of area east and west of the Mississippi River, and contains all Missouri and
Illinois counties highlighted in Figure 1.
St. Louis Metro APA
The most prominent city planning organization within the study area is the St. Louis
Metro section of the Missouri Chapter of the American Planning Association (APA), with 231
member planners as of 2020. This organization covers a substantial portion of the study area,
with a few exceptions. Urban planners within all Missouri counties in the MSA are represented,
but three outlying Illinois counties (Bond, Macoupin, and Calhoun) are not represented by the St.
Louis Metro Section of American Planners Association. Therefore, when local APA mailing lists
were used to solicit survey responses, it can be presumed that planners from Bond, Calhoun, and
Macoupin counties communities were not reached (“St. Louis Metro Section,” n.d.).
Study Subjects – Optimal Candidates
The study subjects of this thesis are active urban planners working at local (or municipal)
planning agencies within the St. Louis MSA. An optimal survey candidate in this study is defined
as a professional staff member (excluding elected / appointed officials) within a municipal
government organization whose job can be clearly identified as including planning and / or
zoning responsibilities, according to any available resources online. There are 316 municipalities
(census
“places”) within the St. Louis MSA (US Census Bureau, n.d.-a). In theory, each of these could
produce an optimal candidate for completion of the survey, however, not every census place has
a professional planner on staff. Many rely on planning boards or commissions of elected or
29
appointed officials to prepare all reports, administrate zoning, and divide up any other typical
planner responsibilities.
A thorough review of web resources for each municipality was conducted to 1) identify
whether there was an optimal candidate within the organization, 2) record their name and job
title, and 3) identify the most direct email address to contact. Optimal candidates identified in this
way were included in an Excel database with their job title and most direct email address, but this
database was exclusively used for participant recruitment and its contents are confidential. This
database shall hereafter be referred to as the Direct-Contact Database. Beginning with a column
of all 316 census places in the study area, optimal candidates were searched for using search
engines. Candidates were sought by combing through the municipality’s webpage, applicable
staff directories, and in some instances, searching the municipality’s social media pages. This
search was run top-to-bottom in alphabetic order to reduce any search bias and resulted in the
identification of optimal candidates at 87 of the 316 municipal organizations. For this reason, the
total population size of optimal candidate organizations in the St. Louis MSA is 87
municipalities. Participant Recruitment
Upon finalization of the survey questions and format within the Qualtrics survey software,
surveys were distributed either through direct solicitation via email or through administrators
from the St. Louis Metro Section of the Missouri Chapter of the American Planning Association.
To solicit survey responses via email directly, the Direct-Contact Database was uploaded to
Qualtrics’ distribution platform after the survey design was finalized. Qualtrics allows the
creation of a short script, which can send an entire table an email introducing the survey and
providing recipients with a hyperlink to partake in the online survey. Where possible, the name
and job title of the optimal candidate were included in the solicitation email to provide the best
30
chance of getting the candidate’s attention in their inbox. Emails sent this way through the
Qualtrics platform track respondents’ participation in the survey, and reminder emails can be sent
to only those who have not opened the survey link or finished their survey submission.
Reminders were sent approximately two weeks and four weeks after the initial invitation. Email
solicitations included a request that only one survey be completed per municipal planning
organization. Following the four-week Qualtrics reminder, administrators of the St. Louis Metro
Section of the Missouri Chapter of the American Planning Association were contacted and asked
to share a short solicitation message (which included a link to the survey) with their mailing list
of active members to boost participation. The administrators sent out the solicitation email and
included information about the survey in their monthly newsletter.
Participants’ Privacy Considerations
Participants’ individual identities were not solicited or recorded in any way. City
demographic / organization characteristics questions at the beginning of the survey captured only
information about the organization that planners work for and its staff, including staff
composition and education. Individual participant’s answers to the entire set of municipal
demographic / organization characteristics questions are not shared. Figures and reports of the
characteristics data include only descriptive statistics of the entire dataset. Any figures showing
individual response data for each of the respondents are coded into alphabetical identifiers to
anonymize the results and preserve each respondent’s confidentiality.
The survey conducted in this study was completely voluntary and came from a willing pool
of organizations and their associated planning professionals. Respondents were informed - in the
introductory and closing sections of the survey itself – of the ways in which their data would be
made confidential, analyzed, and published when the data analysis stage began. Each participant
31
indicated their consent by clicking to continue the introduction stage before completing any
questions. Clear communication of these standards of data management was intended to
encourage more honest and open participation. This study and the survey used to pursue its
research questions were approved by Southern Illinois University Edwardsville’s Institutional
Review Board (IRB) for study on human subjects (approved August 2023). All IRB protocol
regarding the handling and storage of sensitive data was followed to ensure confidentiality for
study participants and their organizations.
Data Sources
Survey
Surveys were conducted using the Qualtrics online survey service, which allowed for high
question customization. The surveys were structured, consisting of an introductory consent
statement, five major data gathering question sub-sections, and a closing / submission note. The
survey questions and overall survey structure were field-tested by a small group of city planners,
planning professionals, and academics adjacent to the urban planning and RS disciplines. This
vetting process with experienced personnel helped all questions to be more sensitive, direct, and
informed to the target participants.
The choice of deploying a survey in this thesis came in response to the unique challenges
associated with gaining a comprehensive view of the integration of RS with urban planning
processes across the St. Louis MSA. Given the complexity of the potential ways and depth with
which organizations across the study area were integrating RS, surveys provided a way to
standardize the data and ask targeted questions. This study set out to establish a mixed-method
framework for calculating integration characteristics and comparing those integration “scores”
across and between organizations to fill gaps in the literature and build upon the integration
32
barriers framework introduced in Hoalst-Pullen and Patterson (2011). Relationships were
assessed between integration scores and potential variables, which may contribute to those levels
of integration representing adoption of RS into planning processes.
Surveys also afforded anonymity of the city planner respondent. They encouraged
respondents to provide candid and honest insight into their use of RS. While use of RS is not
perceived to be a particularly sensitive topic, the factors which contribute to RS use, and
planner’s opinions surrounding this technology, could be something which respondents would
prefer to keep
confidential.
Survey Structure & Questions
This section explores each of the datasets that were collected via the survey to analyze the
research questions presented in this thesis. The following sections describe the survey structure
and content. Tables containing the survey questions (as they were worded in the Qualtrics survey
distributed to respondents) follow the descriptions of each of the sections.
The following sections were included in the distributed version of the survey:
1. Introduction – The introduction welcomed respondents and informed them of both the nature
and purpose of the survey. This section also provided researcher contact information for
follow-up questions, informed respondents of confidentiality and data handling, and solicited
consent from respondents. This section also asked the respondents to provide the municipal
planning organization they represented in the survey, as well as the Missouri or Illinois
county that their municipality was mostly located in. This information was later used to
gather municipal demographics / organization characteristics data, such as geographic size,
total population, and financial expense data.
33
2. Municipal Demographics / Organization Characteristics – Eight datasets were gathered (via
the Census or municipal online financial records) or collected (via the survey) to use in
correlations against the other data categories. Each of these were selected to explore
characteristics of an organization or the individual municipality that the organization
represents (Table 1).
Table 1. Survey Questions for Municipal and Organization Characteristics
Question # Question / Statement Response Options
1 What is the name of the municipality you are
representing in this survey? Write-in
2 What county does the majority of your
municipality’s area belong to? Write-in
3 What was your organization’s 2022 total
annual budget amount (to the nearest $1,000)?
Organizations’ Fiscal Year 2022 Expenses
Actual were subbed in for responses to this
question after the survey closed
4a # of city planner personnel / staff with
planning & zoning responsibilities Write-in (Numeric)
4b # of planner personnel with AICP certification
4c # of in-house GIS professionals
5 Indicate the highest education level for
working planners in your organization Write-in
6 Indicate the highest education for GIS
personnel in your organization Write-In
3. Tools, Techniques, and Applications – This section of the survey asked questions about the
specific RS techniques and technologies that organizations employ, and their relevance to
typical planning tasks outlined in the literature review. The aim of this section was to identify
trends or patterns across the entire respondent dataset on the popular software, techniques,
and their direct relevance to the planning process.
The first portion asked about RS in broad terms, specifically asking about which popular
RS data platforms organizations employed. This section of the survey then asked how often
the planners in respondent organizations use each data product, with options ranging from
‘daily’ to ‘a few times a year.’ The survey then shifts to asking about the value of each RS
34
data product to the completion of specific regular planning tasks, which enabled this survey
to capture data at multiple grains (Table 2).
The final question in this section asked respondents about the software and online
resources they may use to view or manipulate RS data. The question was formatted as ‘Select
all that apply’ with multiple software platforms and online resources / RS data directories
listed. The possible response options were found to be somewhat common through surveying
the popular literature on RS applications in urban planning, and this question did include an
‘other’ option, where participants could write in an answer if an applicable tool or data
directory was not
listed.
Table 2. Survey Questions for Tools, Techniques, and Applications
Question # Question / Statement Response Options
7
Which of the following remote sensing
data types/products does your organization
use?
Aerial Imagery / Photography, Satellite Imagery,
Light Detection and Ranging (LiDAR), Drone
Imagery, Other (Please type answer), My
organization does not use remote sensing
8 How often do planners in your
organization use each data type/product?
A few times a year, A few times a month, Once a
week, A few times a week, Daily
9a How important is each data type/product
for drafting the comprehensive plan?
Not at all important, Slightly Important, Moderately
Important, Very Important, Extremely Important
9b
How important is each data type/product
for creating staff reports for commissions,
committees, or councils?
9c How important is each data type/product
for reviewing permit applications?
9d How important is each data type/product
for creating maps for public use?
9e
How important is each data type/product
for code/zoning/property maintenance
enforcement?
35
10
Which software/online resource does your
organization use to view/manipulate
remote sensing data?
ESRI ArcMap, ESRI ArcPro, EarthExplorer by
USGS, Google Maps, Google Earth, Google Earth
Engine, ERDAS Imagine, QGIS, County GIS
Portal/Viewer, State GIS Portal/Viewer, Other
(Please Specify)
4. Remote Sensing Integration Characteristics – This survey section asked sixteen total
questions about respondent organizations’ integration of RS and urban planning in four
integration areas: financial, technical, applications, and expertise. Four Likert-type questions
were provided for each category, asking questions designed to capture distinct aspects of the
integration. For example, the finance sub-integration section asked a respondent to not only
report whether RS is a funding priority, but also whether current technical spending is
adequate, and if the organization provides a higher compensation to employees with RS
experience. In this way, these sub-integration scores aim to capture the potentially interesting
underpinnings of an organization’s overall score and provide a holistic and nuanced picture
of an organization’s overall RS integration.
The participant was then asked to indicate their agreement to the statements designed to
capture their level of integration with RS in key integration categories identified by Hoalst-
Pullen and Patterson (2011) (Table 3).
Table 3. Remote Sensing Integration Characteristics Survey Questions
Related
F.A.T.E.
Category
Question # Question / Statement Response
Options
11a Remote sensing (technology/data/training) is a priority in my
department’s budgeting discussions
Strongly
Disagree,
Disagree,
Neutral,
Agree,
Strongly
Agree
11b My organization’s spending on remote sensing is adequate to meet
planner’s demands for its use
11c My organization would fund training for planners to learn more
about remote sensing
11d My organization would increase compensation for a planner having
remote sensing experience
12a
Planners in my organization believe remote sensing (raster, pixel,
point-cloud) data is just as valuable as vector data (shapefule, point,
polygon, line) in planning
Financial
36
12b My organization has the technological (computer hardware)
capabilities to support remote sensing analysis for planning
12c My organization often shares remote sensing data with other
municipal / government organizations
12d
My organizations has the software (Arc Pro/Erdas
Imagine/QGIS/Google Earth Engine, etc) capabilities to support
remote sensing analysis for planning
13a Remote sensing helps planning in my organization create/draft plans
13b Remote sensing helps planners in my organization present the public
with information
13c Remote sensing helps planners in my organization with permit
application review
13d Information from Remote sensing helps our organization make
decisions about city planning
14a Experience/training with remote sensing is something our
organization asks about during job interviews for planner positions
14b Planners in my organization have sufficient training/work experience
to complete remote sensing tasks
14c Remote sensing is a topic covered in my organization’s
training/onboarding for new planners
14d My organization offers/encourages ongoing training/education in
remote sensing for urban planning
5. Planner & Organization Views – The Organization / Planners’ Beliefs section asked
respondents to share their level of agreement with seven statements designed to gain insight
into their beliefs and values surrounding the use of RS in urban planning workflows (Table
4).
These responses were recorded using the agreement array Likert-type format described in the
Likert-type questions section in Data Analysis.
Table 4. Planner / Organization Beliefs Survey Questions
Question # Question / Statement Response Options
15a Planners in my organization understand what the term remote
sensing means.
Strongly Disagree, Disagree,
Neutral, Agree, Strongly
Agree
15b Planners in my organization aim to increase the use of remote
sensing technology in the future.
15c Planners in my organization believe remote sensing is valuable
to city planning and decision-making processes.
15d Planners in my organization believe that remote sensing is
convoluted/difficult to understand.
15e Planners in my organization believe remote sensing should be
outsourced to consultants or regional planning organizations.
15f Planners in my organization believe that remote sensing is
Technical Applications Expertise
37
worth dedicating financial resources to.
15g Planners in my organizations believe that remote sensing is a
powerful information-gathering tool.
6. Open-Ended Questions – This short section contained either one or two prompts, depending
on responses to previous questions. If a respondent failed to provide a RS data type that their
organization uses for planning or they indicated that they did not use RS, then they were
presented with the following question: “If your organization does not use remote sensing,
what specifically is preventing planners from using it?” All respondents, whether they
reported using RS or not, were presented with the following question: “What else would you
like to share about your organization’s experience with remote sensing and urban planning?”
Both prompts included text boxes for open-ended responses (Table 5).
Table 5. Open-Ended Survey Section Questions
Question # Question / Statement Response Options
16 If your organization does not use remote sensing, what
specifically is preventing planners from using it? Write-in
17 What else would you like to share about your organization's
experience with remote sensing and urban planning? Write-in
7. Closing & Submission – This section reminded respondents of the confidential nature of
their responses, outlined a general timeline for the completion of the study, provided contact
information for any questions or concerns, and thanked the respondents for their time and
contribution to the thesis.
2020 Census & TIGER Line Shapefile
This study employed data sourced from the 2020 United States Census to capture the 2020
population of each municipality that respondents represented in the survey (US Census Bureau,
n.d.-a). In addition, Census TIGER / Line Shapefiles were used to determine the total 2020
38
geographic area of each municipality in square miles (US Census Bureau, n.d.-b). Both datasets
contribute to the set of municipal demographics / organization characteristics variables that are
used to analyze multiple research questions.
Municipal Budget Reports
One of the organization characteristics survey questions asked for the respondent’s
“organization's 2022 total annual budget amount (to the nearest $1,000),” as a measure of the
municipality’s overall financial capacity. The wording of the question proved problematic due to
its wide range of interpretation by respondents (as either fiscal year 2022 budget allocations,
expenditures, or another figure altogether). To construct a more robust and standardized dataset,
municipal budget reports were obtained from publicly available databases and budget reports
hosted on municipality web resources for respondent organizations. Each budget report provided
the organization’s total expenditures (to the nearest $1,000) for fiscal year 2022, which was
documented and used in the analysis.
Data Analysis
Likert-Type Questions & the ‘Agreement Array’
There is significant debate and controversy around recoding Likert-type response data
into continuous numerical data, and these arguments date back to an article which introduced the
four popular types of data (nominal, ordinal, interval, or ratio), and began discourse on what
forms of statistical analysis may be appropriate for each data type (Stevens 1946). This study
takes the position that Likert-type items can be treated as interval data within Steven’s four-type
data framework, agreeing with one article that, “… subjects often view [Likert-type item scales]
as points on a continuum from low to high with response distributions similar to those obtained
39
on a scaled line with equal intervals between points on that scale” (Willits, Theodori, and Luloff
2016, 9). Through intentionally designing the Likert-item arrays used in quantitative analysis to
mirror the characteristics traditionally agreed upon to make them robust (Mircioiu and Atkinson
2017; Norman 2010; Sullivan and Artino 2013), this study justifies including responses to these
items and arrays in non-parametric bivariate correlation analysis to search for underlying
relationships between municipal demographics / organization characteristics, beliefs, and RS
integration level data sets to investigate the study’s research questions.
The response array for most Likert-type items / sets in this study1 consisted of the
following response options: Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree,
referred to as the Agreement Array. Multiple studies support the inclusion of a neutral middle,
with responses on either end that reflect each other (acts as opposites), as these qualities allow
respondents to assume an equal spacing between options with a neutral middle (Carifio and Perla
2007), which is important to enabling the ability to recode this Likert-type response data.
This study coded the responses (post-hoc) quantitatively from -2 (strongly disagree) to +2
(strongly agree), with 0 acting a neutral middle. An organization responding to an agreement-
style Likert-type question that is recoded to a value of -2 would then translate to a response of
strongly disagree. Averaging all the responses to that Likert-type statement across all responding
organizations could, for example, result in an average value of -1.2. This value would translate to
strong disagreement with the statement provided. This owes to the position taken in this study of
assuming that survey respondents view scales constructed in this way (strongly disagree,
disagree, etc.) as a continuous scale of absolute disagreement to absolute agreement. Building on
this concept, the spectrum of Likert-type response (-2 to +2) is broken up into three agreement
zones: the disagreement zone runs from -2.0 to -0.5, the neutral zone spans -0.5 to +0.5, and the
1 Those located in the Remote Sensing Integration and Planner’s Opinion sections all employ this array.
40
agreement zone spans +0.5 to +2.0. These zones are used to categorize and discuss the findings
from Likert-type questions across the survey response dataset.
As mentioned before, not all the Likert-type items included in the survey and analyzed in
this study were recorded into integer data or treated as anything but ordinal data. There are
Likert-
type items pertaining to Research Question One and included in the Tools, Techniques, and
Applications section of the survey which measured importance of various RS data products to
planning tasks (see Question 9a-9e in Table 2) for the presentation of the Likert-type items. Due
to the nature of the response options lacking a continuous nature with a neutral middle, these
responses were not recoded, and no quantitative analysis was performed on these other than
simple descriptive statistics.
Calculating Integration Scores – The F.A.T.E. Integration Matrix
The initial research interest of this study was whether planners were using RS tools,
concepts, and technologies. To measure a respondent organization’s integration of RS with
planning, a framework was developed called the Financial, Applications, Technical, and
Expertise Integration Matrix (hereafter referred to as the F.A.T.E. Integration Matrix), which is
strongly based on preceding research done by Hoalst-Pullen and Patterson (2011). The F.A.T.E.
Integration Matrix encompasses a set of four integration areas for RS in planning. Hoalst-Pullen
and Patterson (2011) previously referred to these categories as ‘integration barriers,’ but all
barriers can also be presented as opportunities, or areas of improvement. Each component
integration area can be negatively, neutrally, or positively integrated. Much like the recoding of
41
the agreement zones, the analysis of whether a dimension of the F.A.T.E. Integration Matrix is
negatively, neutrally, or positively integrated is developed by averaging the Likert-type responses
to four component Likert-type statements. Each four-Likert-type-statement set is dedicated to
one F.A.T.E. Integration Matrix component, adding up to sixteen total questions that make up the
RS integration portion of the survey (Table 6).
This framework is designed to provide the ability to characterize organizations from their
unique F.A.T.E. RS integration makeup, and then consider the entire study population (in this
case, planners across the St. Louis MSA) by its overall F.A.T.E. characterization. Table 6. F.A.T.E
Integration Matrix Categories and Likert-type Statements
Finance Applications
A Remote sensing (technology / data / training) is a
priority in my department's budgeting discussions. A Remote sensing helps planners in my
organization create / draft plans.
B My organization would fund training for planners to
learn more about remote sensing. B
Remote sensing helps planners in my
organization present the public with
information.
C My organization would fund training for planners to
learn more about remote sensing. C Remote sensing helps planners in my
organization with permit application review.
D My organization would increase compensation for a
planner having remote sensing experience. D
Information from remote sensing helps our
organization make decisions about city
planning.
Technical Expertise
A
Planners in my organization believe remote sensing
(raster, pixel, point-cloud) is just as valuable as vector
data (shapefile, point, polygon, line) in planning.
A
Experience / training with remote sensing is
something our organization asks about during
job interviews for planner positions.
B
My organization has the technological (computer
hardware) capabilities to support remote sensing
analysis for planning.
B
Planners in my organization have sufficient
training / work experience to complete remote
sensing tasks.
C My organization often shares remote sensing data with
other municipal / government organizations. C
Remote sensing is a topic covered in my
organization’s training or onboarding for new
planners.
D My organization has the software capabilities to
support remote sensing analysis for planning. D
My organization offer / encourages ongoing
training / education in remote sensing for urban
planning.
Spearman’s Correlations
Spearman’s correlation coefficients were used to explore whether there existed any
significant (p = 0.05 and p = 0.01) correlations between pairings of municipal demographics /
42
organization characteristics, F.A.T.E. Integration Matrix, and organization beliefs variables.
These pairings included selections from the municipal demographics / organization
characteristics variables, each of the individual integration questions, integration sub-scores,
overall RS integration scores, and planners’ agreement level with each of the questions in the
‘Organization Beliefs / Planner’s Opinions’ section.
Spearman’s rank correlation coefficient (or p) was selected as the bivariate correlation
method due to its ability to handle non-parametric input data (and small sample size). Literature
supports the ability of Spearman’s correlations to generate statistically similar findings to tests of
Likert-type responses using Pearson’s correlation coefficient, ANOVA, and other parametric tests
(Mircioiu and Atkinson 2017; Norman 2010). Two-tailed Spearman’s correlations were
performed in SPSS. These correlations were analyzed for significance at the 95% and 99%
confidence
intervals.
Descriptive Statistics
Throughout the results and conclusions chapters, descriptive statistics are used for the
following purposes: 1) to describe the characteristics of respondent organizations without
breaching confidentiality by reporting trackable organization-specific information; 2) to describe
datasets, including measures of central tendency, ranges, means, medians, and standard
deviations; and 3) to summarize and report on the popular RS tools and datasets used in urban
planning practice throughout the St. Louis MSA.
43
Qualitative Questions and Analysis
The survey was designed to elicit write-in responses throughout each section, providing
opportunities for respondent planners to add more information and context to their responses or
write in selections not available in the multiple-choice dropdowns.
The municipal demographics / organization characteristics section is ordered first in the
survey. This is intentional, as most of the following survey sections ask questions related to
respondents’ use of RS, and it is entirely possible that respondents indicate that they do not use
RS for urban planning at all. If respondents choose to not select any of the options or write in any
data products or platforms (more on that below), they will skip any sections related to their
current use of RS and be directed to the planners / organization beliefs sections. Any respondents
who indicate that they do not use RS, after completing the beliefs section, are then prompted to
share why they do not use RS in planning. It is critical to understand why respondents may not
use RS, and their characteristics and beliefs may provide some insight into what factors could be
contributing to a lack of usage.
For those organizations that do use RS, the Techniques, Tools, and Applications section of
the survey asks about data products, software, and tools. This section also prompts respondents to
rank the importance of their RS tool selections for completing various planning tasks.
Respondents can write-in selections for RS data products. These responses are designed to carry
forward through the ‘importance to planning tasks’ section, allowing respondents to rank the
value of their write-in choices alongside those options provided by default (Aerial Imagery /
Photography, Satellite Imagery, LiDAR, and Drone Imagery). Planners are also asked to provide
any other RS software or data platforms that they use in their work.
44
In the ‘importance to planning tasks’ section, planners are prompted to write in any other
planning tasks that the use of RS data products may be important to. These responses could then
also receive importance rankings for the provided data products and any write-in data products.
This is valuable because it is unlikely that all potential applications for RS in urban planning
workflows have been accounted for, and leaving room for planners to write in how they use RS
products can provide unanticipated insights.
Two more qualitative write-in sections are presented in the closing of the survey, after
planners / organization beliefs. All respondents are offered the opportunity to provide any
additional details related to their organizations’ experience with RS and urban planning. Those
respondents who reported not using any RS are also offered the opportunity to provide insights
about what may be preventing their organizations’ use of RS.
Each of these written sections provide an opportunity to collect qualitative data that can be
coded, and these responses provide context for both respondents’ provided answers as well as
any context surrounding RS that goes beyond the survey questions.
Synthesizing the Results
The discussion and conclusions sections of this study explore the integration of RS in urban
planning in the St. Louis MSA by synthesizing the quantitative, descriptive, and qualitative
datasets generated by the survey, gathered through examining Census statistics for each
respondent municipality, and provided by municipality’s own web-based annual financial reports.
Synthesizing this information examines evidence from all sources and weaves this evidence
together to develop evidence-based conclusions about the current state of RS use in planning
throughout the study area. The discussion and conclusion sections of this thesis then provide
45
recommendations for future research into this topic and the continued growth of RS use in
planning throughout the St. Louis MSA.
Researcher’s Role and Bias / Field Testing Surveys
I am a researcher and a master’s student with previous experience working as a municipal
planner in the St. Louis MSA. My first-hand experience in the planning field undoubtedly means
that I bring to the table my own biases to the research conducted here, particularly about the
ways in which planning organizations are structured, the ways in which I would expect them to
integrate RS, and the barriers/challenges I expect them to face in that integration (or lack
thereof). It is important for me to acknowledge these biases in the pursuit of reproducible and
actionable methods and conclusions.
My survey questions have been developed in collaboration with the knowledge and
experience of other planning professionals and researchers in the field of study. Data capture
through the form of surveys can be susceptible to bias in choice of and wording used for
questions. By checking my bias, I believe that my position of experience in the planning field has
been an asset to this study due to my familiarity with general planning processes, organization
structures, and professional responsibilities.
The structured survey employed in this study was developed with help and feedback from
three professional planners working in this study area, as well as professors and academics
familiar with planning processes and RS integrations. The overall survey structure, flow, and
question set has been modified through a process of face validity to be more concise, targeted,
and audience appropriate. One example of change that was made through this vetting process
was the movement of the introduction and definition of RS directly before any RS-related
questions, which field testers agreed would assist with planners’ comprehension and shared
46
understanding of RS throughout the rest of the survey experience. Multiple drafts of the survey
were completed and suggested edits were provided. The final version of the survey took testers
an average of fifteen minutes to complete from beginning to submission.
CHAPTER 4: RESULTS
Overview
This chapter is structured by breaking the results of the analysis into sections defined by the
five research questions. These research questions build on one another, first asking about the
tools, techniques, and applications used by planners, then examining planning organizations’
value of RS, then the planning organization’s integration of RS and planning, and then asking
whether either organization characteristics of planning organizations or organizations’ beliefs
bear any
relationship to the current level of integration of RS in planning.
Study Population
Before digging into the results for each research question, it is important to provide
information on the pool of respondents whose data is reported in this section. To preserve
respondents’ confidentiality, descriptive statistics are reported at the dataset level, not the
individual organization level.
Response Rate
This study invited responses (through direct email solicitation) from eighty-seven identified
optimal candidates at municipalities across the St. Louis metropolitan statistical area (MSA). Of
the over three hundred ‘places’ identified by the Census in the St. Louis MSA, only these
eightyseven organizations met all the criteria: 1) organizations with planning and zoning
47
authority, 2) contact information for their planning departments on the internet, and 3) have at
least one staff member with dedicated planning and zoning responsibilities (as opposed to an
elected official, such as a planning committee chair).
Survey responses were also solicited through email mailing lists to the members of the St.
Louis metropolitan chapter of the APA, to reach optimal candidates from any organization not
included in the eighty-seven identified through direct solicitation. The culmination of these two
solicitation methods, prompted also by regular survey follow-up reminders, resulted in usable
responses (defined as including responses to at least two of the main data collection sections)
from twelve municipal planning organizations, one of which reported not using RS at all. While
the following subsections describing the study sample, including the organization that did not use
RS, many of the following results sections will primarily focus on the eleven organizations who
reported using RS. With the 87 organizations containing optimal candidates as the denominator,
the twelve respondent organizations resulted in a response rate of 13.8 percent.
Describing the Study Sample – Municipal Demographics / Organization Characteristics
There was an even split among the 12 respondent organizations between those hailing from
municipalities in Illinois (n = 6) and those from municipalities in Missouri (n = 6). The
geographic size of these municipalities ranged from an approximate (in pursuit of confidentiality)
minimum of 4.5 square miles to an approximate maximum of just over 65 square miles. Fiscal
Year 2022 Actual Expenses (for the entire municipal organization) ranged from approximately
$9,000,000 to just over $100,000,000. Municipal populations ranged from 8,000 residents to
50,000 residents.
48
Describing the Study Sample – Planner & GIS Personnel at RS-Using Organizations
Respondent organizations were asked to describe the size and education level of their
planning and GIS departments. The number of dedicated city planner staff ranged from a single
person with planning responsibilities (n = 4) to planning departments with a maximum of nine
planning personnel (n = 1). The organization that reported no use of RS reported four personnel
with planning / zoning responsibilities.
Three organizations had no planner personnel with AICP certification (including the no-RS
organization), up to a maximum of three planners (100%). Planners had an average of 4.2 years
of college education, with a minimum of zero years (n = 1) and a maximum average of 6 years of
college education per planner personnel (n = 4, also the mode by a large margin).
The number of total dedicated Geographic Information Systems (GIS) personnel ranged
from zero (n = 5) to a maximum of just two dedicated GIS professional personnel (n=1). Of the
organizations indicating they had dedicated GIS personnel (n = 7), these professionals attained an
average of 4.7 years of college education, with a minimum of zero years (n = 1), and a maximum
of 6 years (n = 2). The organization that reported no use of RS did not indicate the presence of
any in-house GIS personnel in their municipal organization.
Research Question 1 Results
RQ1: What are the popular RS techniques, technologies, and applications used across the St.
Louis MSA for urban planning?
Planner’s Use of Popular Remote Sensing Data Products
Planning organizations were asked to report their use of the four most popular forms of RS data
capture platforms / products: Aerial Imagery / Photography (AIP), Satellite Imagery, Light
49
Detection and Ranging (LiDAR), and Drone Imagery. Results by anonymized response code are
presented in Table 7. Overall, respondents reported an average utilization of two products per
organization, with a minimum of one and a maximum of four. Planners were provided with the
option to write in additional data platforms, but none were submitted.
Table 7. Remote Sensing Data Product Use by Organization Response Code
Response Code Aerial Imagery /
Photography
Satellite
Imagery
LIDAR Drone Imagery Total
A x x 2
B x x x 3
C x x 2
D x x x 3
E x 1
F x 1
G x x x x 4
H x x 2
I x 1
J x x 2
K x x 2
Totals 11 7 3 2 23
% of
Respondents
100% 64% 27% 18%
All the eleven RS organizations utilized AIP products. Seven of the eleven (64%) utilized
satellite imagery (SI) products. Three of the eleven (27%) utilized LiDAR products. Only two of
the eleven (18%) utilized products from drone imagery (DI).
The utilization of these four different technology product groups varied widely across the
eleven organizations. Only one organization (9%) utilized all four products. Two of the
organizations (18%) utilized three of the four RS products. Five of the organizations (45%)
utilized two of the RS products. Three of the eleven organizations (27%) utilized only one of the
50
RS products, with that product consistently being AIP. Both the average and median number of
RS products utilized by organizations was two products.
Frequency and Importance of RS Product Use
Respondents were asked about their organization’s frequency of use for each of the data
products, along with the relevance of each data product to five common planner tasks /
responsibilities, based on which products they reported using. This line of questioning was
designed to provide context into the ‘what,’ ‘when’, and ‘why’ of broad RS use for municipal
planning organizations. The use frequency level options ranged from Daily to A Few Times a
Year (Table 8).
Aerial Imagery / Photography
Of the municipal planning organizations in the St. Louis MSA that reported using AIP
(n=11), nine subsequently reported using it daily, with two organizations not reporting frequency
of use at all (Table 8).
Table 8. Frequency Option Summaries for Aerial Imagery / Photography Use
Sample
Size Daily
A Few
Times per
Week
Once per
Week
A Few
Times per
Month
A Few
Times per
Year
No
Response
Aerial Imagery /
Photography n = 11 9 2
When asked about the importance of each product for four common planning tasks /
responsibilities, most planners (85.5%) indicated that use of AIP was either ‘Very Important’ or
‘Extremely Important’ no matter the planning use categories. The strongest, high consensus use
cases appeared to be Creating Staff Reports (no responses falling below ‘very important’) and
Creating Maps for Public Use (only one response falling below ‘very important’) (Table 9).
Table 9. Importance of Aerial Imagery / Photography to Planning Tasks
51
Aerial Imagery / Photography Not at all
important
Slightly
important
Moderately
important
Very
important
Extremely
important
Drafting Comprehensive Plan 1 1 3 6
Creating Staff Reports 2 9
Reviewing Permit Applications 1 1 1 8
Creating Maps for Public Use 1 3 7
Code Enforcement 2 1 5 3
Sum 0 5 3 14 33
Sum Percentages 0% 9% 5.5% 25.5% 60%
Satellite Imagery
Seven of the eleven respondent RS-using organizations (64%) indicated using Satellite
Imagery (SI), with five indicating its use at a frequency of ‘Daily’ and two indicating ‘A Few
Times per Week’ (Table 10).
Table 10. Frequency Summaries for Satellite Imagery Use
Sample
Size Daily
A Few
Times per
Week
Once per
Week
A Few
Times per
Month
A Few
Times per
Year
No
Response
Satellite
Imagery n = 7 5 2
Furthermore, of those that reported using Satellite Imagery, 83% indicated that it was
‘Very Important’ or ‘Extremely Important’ across all the five use cases. Code Enforcement was
the weakest of the five use cases with 42.8% reporting an importance level of moderate or lower
(Table 11). Creating Staff Reports and Drafting the Comprehensive Plan were ranked the highest
among use cases, with staff reports garnering six ‘Extremely Important’ responses out of the total
seven.
Table 11. Importance of Satellite Imagery to Planning Tasks
Satellite Imagery Not at all
important
Slightly
important
Moderately
important
Very
important
Extremely
important
52
Drafting Comprehensive Plan 3 4
Creating Staff Reports 1 6
Reviewing Permit Applications 1 1 5
Creating Maps for Public Use 1 1 5
Code Enforcement 2 1 1 3
Sum 0 4 2 6 23
Sum Percentages 0% 11% 6% 17% 66%
Light Detecting and Ranging (LiDAR)
LiDAR lags in frequency and importance of usage when compared to AIP and SI. Only
three of the eleven respondents (27%) indicated using LiDAR (Table 12). Those three
respondents each reported a different frequency of LiDAR use distributed between ‘Once per
week,’ ‘A few times per month’, and ‘A few times per year’ (Table 12). This is notably less often
than the predominantly daily use of AIP or the weekly use of SI.
Table 12. Frequency Summaries for LiDAR Use
Sample
Size Daily
A Few
Times per
Week
Once per
Week
A Few
Times per
Month
A Few
Times per
Year
No
Response
LiDAR n = 3 1 1 1
Furthermore, LiDAR is perceived as ‘Not at all important’ in nearly half (47%) of total
responses across the four use cases (Table 13). The three organizations which reported using
LiDAR unanimously agreed on their perception that LiDAR is not important for Creating Maps
for Public Use (100%), and nearly unanimously agreed to a similar perception regarding its
importance to Code Enforcement (two ‘Not at all important’’ responses, one ‘Slightly important’
response) (Table 13). LiDAR did, however, score moderately to very important for Drafting the
Comprehensive Plan, and one organization indicated it was ‘Extremely Important’ for Reviewing
Permit Applications (Table 13).
Table 13. Importance of LiDAR to Planning Tasks
53
LiDAR Not at all
important
Slightly
important
Moderately
important
Very
important
Extremely
important
Drafting
Comprehensive Plan 2 1
Creating Staff Reports 1 2
Reviewing Permit
Applications 1 1 1
Creating Maps for
Public Use 3
Code Enforcement 2 1
Sum 7 4 2 1 1
Sum Percentages 47% 27% 13% 7% 7%
Drone Imagery
Only two of the eleven respondents (19%) reported using Drone Imagery (DI) (Table 14).
Regarding frequency of DI use, one organization responded with ‘A few times per month,’ while
the other responded with ‘A few times per year’ (Table 14). This is more on par with the
frequency of use of LiDAR than the frequency of AIP or SI use, which both were used on a
weekly or daily basis.
Table 14. Frequency Summaries for Drone Imagery Use
Sample
Size Daily
A Few
Times per
Week
Once per
Week
A Few
Times per
Month
A Few
Times per
Year
No
Response
Drone Imagery n = 2 1 1
Creating Staff Reports was the only planning use case where Drone Imagery was reported
as having an importance level higher than ‘Moderately important’ (Table 15). Drafting the
Comprehensive Plan was the other use case where Drone Imagery was reported by both
respondents as being ‘Moderately important’ (Table 15). All the use cases received at least one
‘Moderately important’ response, indicating a potential middling importance of Drone Imagery
to planning tasks.
Table 15: Importance of Drone Imagery to Planning Tasks
54
Drone Imagery Not at all
important
Slightly
important
Moderately
important
Very
important
Extremely
important
Drafting
Comprehensive Plan 2
Creating Staff Reports 1 1
Reviewing Permit
Applications 1 1
Creating Maps for
Public Use 1 1
Code Enforcement 1 1
Sum 2 1 6 1 0
Sum Percentages 20% 10% 60% 10% 0%
Planner’s Use of Popular Remote Sensing Software / Data Access Portals
The ‘Vertical Sum’ row provides the total number of respondent organizations who use
each software / data access portal (Table 16). Of the total respondent sample (n=11), the vertical
sums minimum is one respondent organization (9.1%), as is the case for users of Earth Explorer
by USGS or Google Earth Engine (Table 16). Whereas the maximum was 11 respondent
organizations (100%), as is the case for users of County GIS Portal / Viewers. Other prominent /
popular tools were Google Maps (10 users, 90.9%) and Google Earth (8 users, 72.7%) (Table
16). While County GIS Portal / Viewer represented the maximum, the State GIS Portal / Viewer
had only two organizations indicate its use (18.2%) (Table 16).
The ‘Horizontal Sum’ column provides the total software / data access portal use per
organization. The horizontal sums range from a minimum of two software / portals used by a
respondent organization to a maximum of seven used (Table 16). Two of the respondent’s (A &
B) wrote-in other software options that their organizations use - ‘EagleView Pictometry’ and
‘NearMap’.
Table 16. Reported Use of RS Software and Data Access Portals
Response ESRI ESRI Earth Google Google Google County State WriteIn Sum
55
Code ArcMap ArcPro
Explorer
by
USGS Maps Earth
Earth
Engine
GIS
Portal /
Viewer
GIS
Portal /
Viewer
A x x x x x x 6
B x x x x 4
C x x x x x 5
D x x x x x x x 7
E x x x 3
F x x 2
G x x x x x 5
H x x x x x 5
I x x x x 4
J x x x x 4
K x x x x 4
Vertical
Sum
8 6 1 10 8 1 11 2 2 49
% of Total
Respondents
72.7% 54.5% 9.1% 90.9% 72.7% 9.1% 100.0% 18.2% 18.2%
Research Question 2 Results
RQ2: Do planners in municipal organizations across the St. Louis MSA believe that RS is
valuable to the urban planning process?
Introduction and Recoding Results
This section covers responses to Question 15 of the survey, which asked planners to
provide ‘Agreement Array’ responses to seven statements designed to capture their organization’s
beliefs regarding the integration of RS and urban planning in a broad sense. Table 17 provides
56
simple response sums for each option in the ‘Agreement Array’. Table 17. Likert-Type Item Responses
to Question 15 (n = 11)
Code "Planners in my organization …" Strongly
Disagree Disagree Neutral Agree Strongly
Agree
A ... understand what the term remote sensing
means 1 2 1 5 2
B ... aim to increase the use of remote sensing
technology in the future 1 1 1 6 2
C ... believe remote sensing is valuable to the
city planning & decision-making processes 1 0 0 6 4
D
... believe remote sensing should be
outsourced to consultants or regional
planning organizations
4 2 4 1 0
E ... believe that remote sensing is convoluted
/ difficult to use or understand 3 6 2 0 0
F ... believe that remote sensing is worth
dedicating financial resources to 1 0 2 7 1
G ... believe that remote sensing is a powerful
information-gathering tool 1 0 0 7 3
While the distribution of these scores is important, responses to these Likert questions can be
recoded to allow easier analysis and discussion. These responses were recoded to integer values
from -2 (‘Strongly Disagree’) to +2 (‘Strongly Agree’), with a neutral middle (0). Recoding these
responses into integers allows for more in-depth descriptive statistics to be run and enables the
inclusion of these statements in bivariate correlation analysis (See Likert results for Research
Questions 3, 4 and 5).
As the ‘Agreement Array’ allows for recoding from -2 to +2, zones have been proposed in
this range that function as guideposts for making conclusions about the scores. Mean response
values, which range between -2 to -0.5, are in the disagreement zone, while mean response
values from -0.5 to 0.5 are in the neutral zone and mean values from 0.5 to 2 are in the agreement
zone.
57
In this context of statistical analysis for Likert responses with a neutral middle, mean
response scores indicate overall levels of agreement across the entire response dataset from
organizations the report using RS (n=11). Variance and standard deviation suggest a level of
agreement or consensus between the RS-using organizations, which can be compared across
responses to different statements to determine which mean response scores are the most reliable
to base conclusions on.
The one organization that did not use RS indicated strong disagreement with all statements
in the organization beliefs section of the survey. Due to the universal nature of their strong
disagreement response, which was not reflected in the responses by any other organization, it
may be reasonable to assume that these responses indicate an outlier. For this reason, these
responses were excluded from correlations calculated between organization beliefs and other
data.
Overall Organization Beliefs Descriptive Statistics
As seen in Table 18, Organization Beliefs Statement A is the only statement in the neutral
zone with a mean response score of 0.45. Statements B (0.64), C (1.09), F (0.64), and G (1.0) are
considered in the agreement zone, with C and G bearing very positive response means, indicating
a stronger agreement across most of the dataset (Table 9). Statements D (0.82) and E (-1.09) are
in the disagreement zone, with E being especially negative, indicating strong disagreement
across most of the dataset (Table 18).
Standard deviation ranged from a minimum of 0.67 (Statement E) to a maximum of 1.23
(Statement A) (Table 18). This indicates considerably more consensus among respondents on
their general disagreement with Statement E (“Planners in my organization believe that remote
sensing is convoluted / difficult to use or understand”) than on their weaker general agreement
58
with Statement A (“Planners in my organization understand what the term remote sensing
means”). Through analyzing each statement this way, a measure of confidence in the consensus
of the eleven respondent organizations can be estimated in the subsequent discussion and
conclusion chapters. Table 18. Descriptive Statistics for Question 15, Organization Beliefs
Code "Planners in my organization …" Min. Max. Mean Std Deviation
A ... understand what the term remote sensing
means -2 2 0.45 1.23
B ... aim to increase the use of remote sensing
technology in the future -2 2 0.64 1.15
C ... believe remote sensing is valuable to the
city planning & decision-making processes -2 2 1.09 1.08
D
... believe remote sensing should be
outsourced to consultants or regional
planning organizations
-2 1 -0.82 1.03
E ... believe that remote sensing is convoluted
/ difficult to use or understand -2 0 -1.09 0.67
F ... believe that remote sensing is worth
dedicating financial resources to -2 2 0.64 0.98
G ... believe that remote sensing is a powerful
information-gathering tool -2 2 1 1.04
Organization Beliefs A: Understanding Remote Sensing
The statement ‘Organization Beliefs A’ is one designed to determine whether planners and
their organizations feel as though they have a grasp on RS as an overall concept. The full
statement is ‘Planners in my organization understand what the term remote sensing means.’ The
responses to this statement demonstrate the highest variance (1.52) and standard deviation (1.23)
across the entire beliefs section of the survey, indicating that there is notable variation /
dispersion in the responses (Table 18). The response mean ultimately falls at the top of the
neutral zone with a mean of 0.45, just under the 0.5 border with the agreement zone (Table 18).
59
Organization Beliefs B: Increasing Remote Sensing Use
Organization Beliefs B is a statement which is designed to explore planning
organizations’ valuation of RS by whether they are interested in further developing or investing
in RS use. The full statement is ‘Planners in my organization aim to increase the use of remote
sensing technology in the future.’ The variance (1.32) and standard deviation (1.15) of responses
to this statement indicated a notably low consensus across reporting planning organizations
(Table 18). The mean score for this statement was 0.64, which barely fell into the agreement zone
(Table 18).
Organization Beliefs C: Value of Remote Sensing to City Planning
Only one of the eleven respondents disagreed with Organization Beliefs C (‘Planners in my
organization believe remote sensing is valuable to the city planning & decision-making
processes’), and this dataset generated middling consensus metrics (variation = 1.17, std. dev. =
1.08). This resulted in the strongest positive mean (1.09) in the beliefs dataset, one which is
squarely and strongly in the agreement zone (Table 18).
Organization Beliefs D: Outsourcing Remote Sensing
Organization Beliefs D statement is ‘Planners in my organization believe remote sensing
should be outsourced to consultants or regional planning organizations.’ Responses to this
statement provided middling consensus metrics (variation = 1.06, std. dev = 1.03) (Table 18).
This is the first statement in the series which generates a mean response value (-0.82) in the
disagreement zone (Table 18). It should also be noted that of the eleven responses to this
statement, four of them (36%) indicated a neutral stance towards the statement (Table 18). This is
twice as many as any other question.
60
Organization Beliefs E: Remote Sensing Confusion
Organization Beliefs E is a statement designed to explore the respondents’ organizations’
potential frustrations with understanding or using RS. The full statement is ‘Planners in my
organization believe remote sensing is convoluted / difficult to understand.’ This statement
presents the most promising measures of consensus with the lowest variance (0.45) and standard
deviation (0.67) measures of the entire beliefs section of the survey (Table 13). The question also
stands out for being strongly set in the disagreement zone with a mean response value of -1.09
(Table 13).
Organization Beliefs F: Financial Value of Remote Sensing
Organization Beliefs F explores specifically the financial value of RS, or the perceived
return of investing in RS as a technology, capability, and resource. The full statement is
‘Planners in my organization believe that remote sensing is worth dedicating financial resources
to.’ Reponses to this statement had middling consensus metrics (variance = 0.96, std. dev. =
0.98), and a relatively weak mean score of 0.64, placing the mean response score just barely in
the realm of the agreement zone (Table 18).
Organization Beliefs G: Remote Sensing to Gather Information
The final Organization Belief Statement (G) is designed to explore the value of RS, not just
as a concept or capability, but specifically as a data-gathering and knowledge building asset. The
full statement is ‘Planners in my organization believe that remote sensing is a powerful
information-gathering tool.’ This Likert statement showed signs of middling consensus (variance
= 1.09, std. dev. = 1.04), suggesting moderate agreement between respondents (Table 19). The
mean response score of 1.00 suggests an overall response which, much like Organization Beliefs
C, falls squarely in the agreement zone (Table 19). This suggests a strong positive response.
61
Research Question 3 Results
RQ3: What is the current level of RS and planning integration across the St. Louis MSA, based
on the F.A.T.E. Integration Matrix?
Average Remote Sensing Integration & F.A.T.E. Results Overview
A total RS integration score is calculated for each organization by averaging the four
integration sub-section scores from the F.A.T.E Integration Matrix. In this way, the aim was to
create a comprehensive and replicable assessment of each organization’s overall integration of
RS into urban planning activities of the practitioners. Using the same recoding technique for the
‘Agreement Array’, these numbers occupy a possible range from -2 (strongly disagree / low
integration) to +2 (strongly agree / high integration). These integration category scores, and
average overall integration scores, provide a holistic view of the integration characteristics of
each municipal planning organization.
Average Integration scores (Table 19) ranged from 0.375 to 1.31, with a mean average
integration score (0.69) that falls weakly in the positive integration zone. This suggests weak
positive integration of RS for planning across all respondents throughout the St. Louis MSA.
Table 19. Respondents' Integration Categories and Average Integration Scores
Response Code Financial Applications Technical Expertise Average Integration
A 1.750 1.000 1.500 1.000 1.313
B 0.750 1.000 0.250 0.750 0.688
C 0.750 1.250 0.000 -0.250 0.438
D 0.500 1.250 1.000 0.500 0.813
E 0.500 2.000 1.250 0.000 0.938
F 1.000 1.000 0.000 -0.500 0.375
G 1.000 2.000 1.500 0.500 1.250
H 0.750 0.750 1.000 0.000 0.625
62
I -0.250 1.750 -0.250 0.500 0.438
J -1.000 2.000 0.750 -0.250 0.375
K 0.500 1.000 0.250 -0.250 0.375
The Average Integration score has the lowest variance (0.35) and standard deviation (0.12)
(Table 20). Average integration demonstrated a smaller range of scores (0.94) than any of its
F.A.T.E. components, though that is likely due to the nature of its calculation as being a
composite of its component parts (Table 20). Similarly, it demonstrated the lowest variation
(0.35) and standard deviation (0.12) (Table 20).
Table 20. Descriptive Statistics for Integration Scores
Integration Category Minimum Maximum Mean (Average
Integration Score)
Standard
Deviation
Financial -1.00 1.75 0.57 0.50
Applications 0.75 2.00 1.36 0.23
Technical -0.25 1.50 0.66 0.40
Expertise -0.50 1.00 0.18 0.24
Average Integration 0.38 1.31 0.69 0.12
F.A.T.E. Integration Matrix Landscape
The following chart (Figure 3) displays stacked (cumulative / summed) F.A.T.E. Integration
Matrix scores. It also provides, with the white horizontal lines, the respondent organization’s
overall average integration score across all four of the F.A.T.E. categories (Figure 3).
Respondents are ordered from highest (left) average integration to lowest (right) average
integration (Figure 3).
63
As the bars in this graph are stacked to represent summarized F.A.T.E. Integration Matrix
scores, it has a maximum range of -8.0 to +8.0. This range is due to each individual F.A.T.E.
Integration Matrix score having a minimum and maximum possible value of -2.0 to +2.0, and
since there are four of them, that range must be multiplied by four. An organization with a
stacked bar nearing 8, for instance, would have responses that are quite positive in all four
F.A.T.E Integration Matrix categories. Similarly, the average integration score bars in the chart
are related to the average size and directionality (negative or positive) of the bars themselves and
have their own range of -2.0 to +2.0 (Figure 3). Figure 3 is effectively a snapshot of the
distribution of RS integration with urban planning across the St. Louis MSA, based on the
F.A.T.E. Integration
Matrix and the number of organizations that responded to the survey.
Figure 2. Integration Scores by Respondent
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
A G E D B H C I F J K
F.A.T.E. Integation Matrix & Average Integration Scores by Respondent Code
Financial Applications Technical Expertise Average Integration
64
Responses from all eleven organizations demonstrate overall integration levels above zero.
With a few exceptions, the applications bar for each organization is the largest. Visual analysis
suggests that even when the cumulative bar shrinks (as overall integration decreases from left to
right), the applications section still appears to maintain size relative to the other sections (Figure
3). Applications also never go negative, while the expertise integration bar is negative in over a
third of the cases. The only bars which ever disappear, suggesting true neutrality per the
Likerttype item scales, are technical integration (n = 2) and expertise integration (n = 1). All the
organizations that present overall integration in the positive integration zone maintain either
neutral or positive integration directions for all four F.A.T.E. Integration Matrix categories
(Figure 3). Only those organizations with average integration in the neutral integration zone
(average integration of -0.5 to 0.5) have any F.A.T.E. Integration Matrix categories presenting in
the negative direction, with those most often being expertise integration (n = 4) and financial
integration (n = 2).
Financial Integration Descriptive Statistics
Each of the statements from the Financial Integration portion of the survey targeted a
different aspect of an organization’s willingness to fund the development of RS within their
planning practice. The summary of Likert selections is displayed in Table 21 below.
65
Table
21. Likert Response Sums for Each Financial Integration Statement
Code Financial Integration Statement
Strongly
Disagree (-
2)
Disagree (-
1)
Neutral
(0) Agree (1) Strongly
Agree (2)
A
Remote sensing (technology / data /
training) is a priority in my
department’s budgeting discussions
1 5 2 3
B
My organization’s spending on remote
sensing is adequate to meet planner’s
demand for its use.
2 1 6 2
C
My organization would fund training
for planners to learn more about
remote sensing
2
7 2
D
My organization would increase
compensation for a planner having
remote sensing experience
2 7 1 1
Out of the full -2 to +2 score range, the financial integration score across all respondent
organizations had a mean of 0.57, which is a score on the very low end of the positive integration
zone (Table 20 in Average Remote Sensing Integration & F.A.T.E. Results Overview, above). In
the full range of integration scores (range = +0.18 to +1.36), this was the second most positively
integrated F.A.T.E. category of the four (Table 20). The financial integration dataset also
demonstrated the highest variability measures of the four categories (range = 2.75, variance =
0.50, std. dev. = 0.71) (Table 14).
The financial integration average across all participants is pulled down by the financial
integration statement ‘My organization would increase compensation for a planner having
remote sensing experience’ (Finance Statement A, mean = 0.09) (Table 22). That statement had
the lowest variability measures (variance = 0.63, std. dev. = 0.79) of the four financial
statements, suggesting a strong consensus (Table 22). Statements A (mean = 0.64), B (mean =
66
Table
0.73), and C (mean = 0.82) all presented in the weak agreement zone (Table 22). These
statements related to budget priorities, spending adequacy, and funding training.
22. Descriptive Statistics for Financial Integration Likert Responses
Code Financial Integration Statement Minimum Maximum Mean Std
Deviation
A
Remote sensing (technology / data /
training) is a priority in my department’s
budgeting discussions
-1 2 0.64 0.98
B
My organization’s spending on remote
sensing is adequate to meet planner’s
demand for its use.
-1 2 0.73 0.96
C
My organization would fund training for
planners to learn more about remote
sensing
-1 2 0.82 0.94
D
My organization would increase
compensation for a planner having
remote sensing experience
-1 2 0.09 0.79
Applications Integration Descriptive Statistics
Applications integration Likert questions asked organizations to measure the value of RS as
a solution to various business problems / use cases that they experience. Each of the four
questions in this section explored a different area of potential applications for RS in planning.
This table includes a sample size column, as applications Statement C was the only F.A.T.E.
Integration Matrix component question that one of the eleven respondents skipped (Table 23).
None of the respondents reported disagreeing with any of the four applications statements across
the entire response dataset (Table 23).
23. Likert Response Sums for Each Applications Integration Statements
Code Applications Integration
Statement
Strongly
Disagree
(-2)
Disagree
(-1)
Neutral
(0)
Agree
(1)
Strongly
Agree (2)
Sample
Size (n)
A RS helps planners in my 1 7 3 11
67
Table
organization create / draft
plans.
B
RS helps planners in my
organization present the public
with information.
1 4 6 11
C
RS helps planners in my
organization with permit
application review.
6 4 10
D
Information from RS helps our
organization make decisions
about city planning.
6 5 11
Overall, the mean applications integration score across all the respondent organizations was
1.36 (see Table 20 in Average Remote Sensing Integration & F.A.T.E. Results Overview). This is
strongly in the positive integration zone, and the relatively low variance (0.23) and standard
deviation (0.48) suggests a strong consensus among the respondents (Table 20). Average
responses to all four of the Likert statements composing the applications integration section of
the survey fell strongly in the agreement zone (Table 24).
The weakest Likert statement (Statement A: ‘Remote sensing helps planners in my
organization create / draft plans’) had a mean score of 1.18 and had no disagreement responses
(Table 24). Of the planning tasks presented to them, planners believe that RS aids them most
with presenting information to the public (Statement B, mean = 1.45) and making decisions
about city planning (Statement D, mean = 1.45), closely followed by reviewing permit
applications
(Statement C, mean = 1.40) and creating and drafting plans (Statement A, covered above) (Table
24). It is important to note, however, that applications Statement C has a slightly smaller sample
size (n=10) due to one participant skipping it.
68
Table
24. Descriptive Statistics for Applications Integration Likert Responses
Code Applications Integration Statement Minimum Maximum Mean Std
Deviation
A Remote sensing helps planners in my
organization create / draft plans. 0 2 1.18 0.57
B
Remote sensing helps planners in my
organization present the public with
information.
0 2 1.45 0.66
C
Remote sensing helps planners in my
organization with permit application
review.
1 2 1.4 0.49
D
Information from remote sensing helps
our organization make decisions about
city planning.
1 2 1.45 0.50
Technical Integration Descriptive Statistics
The technical integration sub-section of the survey asked respondents to indicate their
agreement (via Likert-scale questions) to four technical capacity-related integration statements.
Each of the technical integration Likert statements sought information on an organization’s value
in the technology and technical capabilities needed to utilize remote sensing data for planning
practice.
Technical integration questions measure an organization’s technical capacity for handling
RS data, their beliefs towards RS data types, and their willingness to share RS data with other
organizations. The mean technical integration score across all eleven respondents was 0.66 (see
Table 20 in Average Remote Sensing Integration & F.A.T.E. Results Overview). This places
technical integration in the lower positive integration zone. This dataset had the second-highest
variance (0.40) and standard deviation (0.64) (Table 20), suggesting it was more polarizing for
respondents than applications or expertise integration areas.
25. Likert Response Sums for Each Technical Integration Statements
69
Table
Code Technical Integration Statement
Strongly
Disagree
(-2)
Disagree
(-1)
Neutral
(0)
Agree
(1)
Strongly
Agree (2)
A
Planners in my organization believe remote
sensing (raster, pixel, point-cloud) is just as
valuable as vector data
(shapefile, point, polygon, line) in planning
5 4 2
B
My organization has the technological
(computer hardware) capabilities to
support remote sensing analysis for
planning
1 3 3 4
C
My organization often shares remote
sensing data with other municipal /
government organizations 1 3 3 4
D
My organization has the software (Arc Pro
/ Erdas Imagine / QGIS / Google Earth
Engine, etc) capabilities to support remote
sensing analysis for planning
2 6 3
One technical integration statement means response (Statement D, mean = 1.09) fell strongly
in the agreement zone (Table 26). Two of the mean responses for technical integration statements
(Statement A, mean = 0.73 and Statement B, mean = 0.91) fell weakly into the agreement zone
(Table 26). The overall technical integration score would have been stronger were it not for
responses to Statement C, ‘My organization often shares remote sensing data with other
municipal / government organizations,’ which has a response mean of -0.73 (Table 26).
Responses to this statement demonstrated an astronomical amount of variance (7.47) and
standard deviation (2.73) compared all other integration statements in the entire F.A.T.E.
Integration Matrix, suggesting that this statement was quite polarizing relative to other
statements (Table 26).
26. Descriptive Statistics for Technical Integration Likert Responses
Code Technical Integration Statement Minimum Maximum Mean Std
Deviation
70
Table
A
Planners in my organization believe
remote sensing (raster, pixel, point-cloud)
is just as valuable as vector data
(shapefile, point, polygon, line) in
planning
0 2 0.73 0.75
B
My organization has the technological
(computer hardware) capabilities to
support remote sensing analysis for
planning
-1 2 0.91 1.00
C
My organization often shares remote
sensing data with other municipal /
government organizations
-2 1 -0.73 2.73
D
My organization has the software (Arc
Pro / Erdas Imagine / QGIS / Google
Earth Engine, etc) capabilities to support
remote sensing analysis for planning
0 2 1.09 0.67
Expertise Integration Descriptive Statistics
The expertise integration sub-section of the survey asked respondents to indicate their
agreement (via Likert-scale questions) to four expertise capacity-related integration statements.
Each of these statements sought information on an organization’s current level of institutional
knowledge surrounding RS use and their commitment to expanding or maintaining that
knowledge base.
27. Likert Response Sums for Each Expertise Integration Statements
Code Expertise Integration Statement
Strongly
Disagree (-
2)
Disagree (-
1)
Neutral
(0)
Agree
(1)
Strongly
Agree (2)
A
Experience / training with remote sensing is
something our organization asks about during
job interviews for planner positions.
3 4 4
B
Planners in my organization have sufficient
training / work experience to complete remote
sensing tasks.
1 2 8
C
Remote sensing is a topic covered in my
organization's training or onboarding for new
planners.
1 5 4 1
71
Table
D
My organization offers / encourages ongoing
training / education in remote sensing for urban
planning
2 2 6 1
Expertise integration suffered from the lowest integration of the four F.A.T.E. Integration
Matrix categories with a mean score of 0.18, placing it squarely in the neutral integration zone
(see Table 14 in Average Remote Sensing Integration & F.A.T.E. Results Overview). Overall
expertise integration also demonstrated middling consensus metrics, with a standard deviation of
0.24 (Table 20).
Two of the expertise statement responses presented mean response values in the agreement
zone: Statement B (mean = 0.64) ‘Planners in my organization have sufficient training / work
experience to complete remote sensing tasks,’ and Statement D (mean = 0.55) ‘My organization
offers / encourages ongoing training / education in remote sensing for urban planning’ (Table
28).
Responses to Statement A (mean = 0.09), ‘Experience / training with remote sensing is
something our organization asks about during job interviews for planner positions,’ fell into the
neutral zone
(Table 28). The mean response to Statement C, ‘Remote sensing is a topic covered in my
organization's training or onboarding for new planners’ was -0.55 (Table 28). This statement
was designed to understand awareness of RS and its value for planning within the planning
72
organization more broadly. This response presented relatively strong consensus metrics (variance
= 0.61, std. dev. = 0.78) (Table 28).
Table 28. Descriptive Statistics for Expertise Integration Likert Responses
Code Expertise Integration Statement Min. Max. Mean Std Deviation
A
Experience / training with remote
sensing is something our organization
asks about during job interviews for
planner positions.
-1 1 0.09 0.79
B
Planners in my organization have
sufficient training / work experience to
complete remote sensing tasks.
-1 1 0.64 0.64
C
Remote sensing is a topic covered in my
organization's training or onboarding
for new planners.
-2 1 -0.55 0.78
D
My organization offers / encourages
ongoing training / education in remote
sensing for urban planning
-1 2 0.55 0.89
Research Question 4 Results
RQ4: Do demographic qualities of a municipality or characteristics of the municipal planning
organization have a measurable effect on an organization’s level of RS integration in urban
planning?
This section now transitions to exploring the results of the Spearman’s Correlation analysis
used to search for significant relationships between demographic qualities of municipal planning
organizations (and the cities they represent) and organization’s F.A.T.E. Integration Matrix
levels, or average RS integration (Table 29). This study also included each individual F.A.T.E.
Integration Matrix component question (four per Matrix integration category) in the Spearman
Correlation analysis against the municipal demographics / organization characteristics variables
73
to search for what specific aspects of an area of integration bear a strong relationship with
demographic variables. This study used five organization characteristics of planning
organizations and their
cities:
•Number of Planner Personnel – Total number of personnel within the organization with
active planning responsibilities.
•Number of Geospatial Information Systems (GIS) Personnel – Total number of active
GIS professionals on staff.
•Percent of Planners with AICP - Percentage of planning staff who are certified by the
American Institute of Certified Planners (AICP).
•Average. Planners Education (Yrs.) – The average post-high school years of education
attained by planning personnel.
•Average GIS Personnel Education (Yrs.) – The average post-high school years of
education attained by GIS personnel.
Financial Integration / Organization Characteristics Correlations
Organizations’ financial integration score demonstrated a significant negative relationship
with their total number of GIS personnel (95% C.I., rs = -0.630, p = 0.038). Another result in this
section is that of finance integration Likert statement B: ‘My organization would fund training
for planners to learn more about remote sensing’ demonstrated strong positive Spearman’s
correlations with two organization characteristics: Number of planner personnel (95% C.I., rs =
75
Table 29. Spearman’s Coefficients - Integration Scores and Organization Characteristics
* = Sig. at 95% C.I.
** = Sig. at 99% C.I.
2020
Geographic Size
FY 2022
Expenditures
2020
Census
Population
# of
Planner
Personnel
# of GIS
Personnel
% of
Planners
w/
A.I.C.P.
Avg. Yrs.
Planners'
Education
Avg. Yrs.
GIS
Personnel
Education
Sample
Size (n)
Statement A -0.382 -0.014 -0.266 0.376 0.364 -0.189 -0.440 0.082 11
Statement B 0.305 -0.184 0.404 .680* .678* 0.090 0.097 .746** 11
Statement C -0.119 0.143 0.000 0.514 0.564 -0.368 -0.146 0.506 11
Statement D -0.079 -0.037 -0.296 0.282 0.250 -.630* -0.359 0.269 11
Finance Score -0.088 0.135 0.019 0.598 .624* -0.196 -0.297 0.419 11
Statement A 0.401 0.267 0.107 -0.230 0.361 .773** 0.245 0.277 11
Statement B 0.562 -0.076 0.433 0.052 0.315 0.590 0.183 0.352 11
Statement C -0.107 -0.071 -0.355 -.650* -0.077 .883** 0.220 -0.112 10
Statement D -0.318 -0.346 -0.462 -0.563 -0.217 0.564 -0.059 -0.215 11
App. Score 0.418 0.009 0.207 -0.181 0.281 .777** 0.217 0.268 11
Statement A -0.162 -0.191 -0.544 -0.423 0.055 0.461 0.090 0.211 11
Statement B 0.231 -0.294 0.009 .667* 0.528 -0.327 -.748** 0.373 11
Statement C 0.126 0.090 -0.033 -0.005 0.339 0.368 -0.260 0.169 11
Statement D 0.078 -0.302 -0.357 0.309 0.235 -0.258 -.797** 0.133 11
Technical Score 0.113 -0.179 -0.298 0.228 0.449 0.040 -0.582 0.324 11
Statement A 0.010 0.241 -0.294 -0.250 -0.114 0.414 -0.076 -0.315 11
Statement B 0.268 0.041 0.534 0.539 .721* 0.409 0.315 .610* 11
Statement C -0.203 -0.088 0.005 0.186 -0.200 -0.219 -0.210 -0.348 11
Statement D 0.202 -0.154 0.070 0.476 0.236 -0.241 -0.286 0.175 11
Expertise Score 0.067 -0.116 0.009 0.312 0.187 0.115 -0.188 -0.002 11
Average Integration Score 0.152 -0.308 -0.060 0.455 .606* 0.130 -0.351 0.481 11
Sample Size (n) 11 11 11 11 11 11 11 11
Finance Applications Technical Expertise
76
0.680, p = 0.021) and number of GIS personnel (95% C.I., rs = 0.678, p = 0.021) (Table 29).
The strongest correlation in the financial integration realm was between financial integration
Statement C (‘My organization would fund training for planners to learn more about remote
sensing’’) and an organization’s GIS personnel’s average years of college education (99% C.I., rs
= 0.746, p = 0.008) (Table 29).
Applications Integration / Organization Characteristics Correlations
The applications integration seems to be related to two characteristics: organizations’ total
number of planner personnel and organizations’ total percentage of planners who hold an
American Institute of Certified Planners (AICP) certification. As for the former, total number of
planner personnel demonstrated a strong positive Spearman’s correlation (95% C.I., rs = 0.650, p
= 0.042) with application Likert statement C (‘Remote sensing helps planners in my organization
with permit application review’) (Table 29). Percentage of planners with AICP certificate
demonstrated especially strong positive Spearman’s correlations with two application Likert
questions: Statement A (‘Remote sensing helps planners in my organization create / draft plans,’
99% C.I., rs = 0.773, p = 0.005) and Statement B (‘Remote sensing helps planners in my
organization with permit application review,’ 99% C.I., rs = 0.883, p = 0.001) (Table 29).
Technical Integration / Organization Characteristics Correlations
Two technical integration Likert statements stood out in the Spearman’s Correlation
analysis. Technical B (‘My organization has the technological (computer hardware) capabilities
to support remote sensing analysis for planning’) achieved both a significant positive correlation
(95% C.I., rs = 0.667, p = 0.025) with number of planning personnel and a very significant
77
negative correlation (99% C.I., rs = -0.748, p = 0.008) with average years of planner education
(Table 29).
That same characteristic (years of planner education) shared a very significant negative
correlation
(99% C.I., rs = -0.797, p = 0.003) with Technical D (‘My organization has the software
capabilities to support remote sensing analysis for planning’) (Table 29).
Expertise Integration / Organization Characteristics Correlations
While no expertise integration statements resulted in very significant correlations (99%
confidence interval), two statements did achieve significance at the 95% confidence interval.
These were both achieved by expertise integration Likert statement B (‘Planners in my
organization have sufficient training / work experience to complete remote sensing tasks’), which
was positively correlated with an organization’s total number of GIS personnel (rs = 0.721, p =
0.012) and organizations’ average years of GIS personnel education (rs =0.610, p = 0.046) (Table
29).
Average integration Score & Organization Characteristics Correlations
The organization’s overall RS integration score, averaged across the four F.A.T.E.
Integration Matrix component scores, showed a significant Spearman’s correlation (95% C.I., rs
= 0.606, p = 0.048) with their total number of GIS personnel (Table 29).
Research Question 5 Results
RQ5: Do either organization characteristics or integration scores have a significant relationship
with organization opinions on the integration of RS and urban planning?
78
Planner’s Opinion Questions
Survey questions 15A through 15G captured the responding organizations’ views on RS
within their sphere of planning operations. These questions are referred to interchangeably as
Organization Beliefs and Question 15 Responses.
All seven of the Likert statements presented in the planner’s / organization opinions
section of the survey generated significant Spearman’s correlations against either the F.A.T.E.
Integration Matrix categories or organizations’ total average RS integration scores (see Table 30).
None of the seven ‘Organization Beliefs’ statements (coded A through G) generated any
significant correlations when placed against the municipal demographics / organization
characteristics variables presented in the previous section. This result, therefore, suggests that
organization characteristics do not significantly impact planners’ opinions on the integration of
RS and urban planning.
Organization Beliefs A
The first planner / organization beliefs Likert statement is ‘Planners in my organization
understand what the term remote sensing means,’ and it presented a significant positive
correlation with expertise integration Likert Statement A (‘Experience / training with remote
sensing is something our organization asks about during job interviews for planner positions’).
Organization Beliefs B
The second Likert statement in this section (‘Planners in my organization aim to increase
the use of remote sensing technology in the future’) were found to have a significant positive
79
correlation with two F.A.T.E. Integration Matrix component Likert statements: Technical A
(‘Planners in my organization believe remote sensing (raster, pixel, point-cloud) is just as
valuable as vector data (shapefile, point, polygon, line) in planning’, 95% C.I., rs = 0.646, p =
0.044), and Expertise A (‘Experience / training with remote sensing is something our
organization asks about during job interviews for planner positions’, 95% C.I., rs = 0.755, p =
0.012) (Table 30).
Organization Beliefs C
Organization / planner beliefs Likert statement C (‘Planners in my organization believe
remote sensing is valuable to the city planning & decision-making processes’) generated a single
significant positive correlation (95% C.I., rs = 0.640, p = 0.046) with expertise integration Likert
statement A (‘Experience / training with remote sensing is something our organization asks about
during job interviews for planner positions’) (Table 30).
Organization Beliefs D
This statement: ‘Planners in my organization believe remote sensing should be
outsourced to consultants or regional planning organizations’ generated a single very significant
negative correlation (99% C.I. rs = -0.779, p = 0.008) with financial integration Likert statement
C (‘My organization would fund training for planners to learn more about remote sensing’)
(Table 30).
80
Table 30. Spearman’s Coefficients - Integration Scores and Organization Beliefs
* = Sig. at 95% C.I. **
= Sig. at 99% C.I.
Statement
A Statement B Statement C Statement D Statement E Statement F Statement G Sample
Size (n)
Statement A 0.080 0.147 0.153 -0.512 -0.118 0.475 0.530 11
Statement B 0.266 -0.280 0.000 -0.462 -0.500 -0.034 0.345 11
Statement C -0.177 0.000 -0.323 -.779** 0.000 0.578 0.345 11
Statement D -0.324 -0.031 -0.362 -0.421 0.062 0.542 0.301 11
Finance Score 0.060 0.010 -0.036 -0.554 -0.084 0.446 0.464 11
Statement A 0.495 .646* 0.533 -0.095 -0.589 0.401 .732* 11
Statement B 0.088 -0.149 0.000 -0.350 -0.514 0.498 0.592 11
Statement C 0.113 0.313 0.261 -0.140 -0.318 0.495 0.478 10
Statement D 0.384 0.249 0.323 0.000 -0.500 0.578 .690* 11
App. Score 0.296 0.363 0.287 -0.250 -0.583 .752* .805** 11
Statement A 0.408 0.344 0.356 0.239 -0.345 0.141 0.524 11
Statement B 0.153 -0.321 0.250 0.447 -0.323 -0.527 0.089 11
Statement C 0.334 0.334 0.550 0.450 -0.207 -0.207 0.316 11
Statement D 0.420 0.321 0.583 0.373 -0.323 -0.220 0.356 11
Technical Score 0.331 0.029 0.410 0.420 -0.375 -0.295 0.319 11
Statement A .683* .755* .640* 0.195 -0.379 0.476 0.443 11
Statement B 0.074 -0.352 -0.051 -0.567 -0.432 -0.107 0.325 11
Statement C -0.077 0.126 0.154 -0.303 -0.238 0.308 -0.041 11
Statement D 0.070 -0.093 0.201 -0.072 -0.311 0.152 0.258 11
Expertise Score 0.334 0.197 0.364 -0.241 -0.621 0.346 0.428 11
Average Integration Score 0.304 0.045 0.253 -0.439 -.783** 0.411 .810** 11
Sample Size (n) 10 10 10 10 10 10 10
Finance Applications Technical Expertise
Organization Beliefs E
The fifth organization / planner beliefs Likert statement, ‘Planners in my
organization believe that remote sensing is convoluted / difficult to use or understand’
was found to have a very significant negative correlation (99% C.I., r = -0.783, p =
0.008) with organizations average RS integration score (Table 30).
Organization Beliefs F
The statement, ‘Planners in my organization believe that remote sensing is
worth dedicating financial resources to,’ had a significantly positive Spearman’s
correlation (95% C.I., rs = 0.752, p = 0.012) with organizations’ overall technical
integration score (Table 30).
Organization Beliefs G
The statement, ‘Planners in my organization believe that remote sensing is a
powerful information-gathering tool,’ had a significant positive correlation with two
technical statements.
One very significant positive correlation with organizations’ overall technical
integration score and one very significant positive correlation with organizations’
overall average integration score across the four F.A.T.E. Integration Matrix categories.
The two technical statements that Organization Beliefs G was significantly
correlated with are Technical A (‘Planners in my organization believe remote sensing
(raster, pixel, point-cloud) is just as valuable as vector data (shapefile, point, polygon,
line) in planning,’ 95% C.I., rs = 0.732, p = 0.016) and Technical D (‘My organization
has the software capabilities to support remote sensing analysis for planning,’ 95%
C.I., rs = 0.690, p = 0.027) (Table 30). Furthermore, the
Organization Beliefs G generated a very significant positive correlation with overall
technical integration (99% C.I., rs = 0.810, p = 0.004) and a very significant positive
correlation with overall average RS integration score (99% C.I., rs = 0.805, p = 0.005)
(Table 30).
Open-Ended Questions:
Three statements were provided by respondents for the open-ended questions
(16 and 17) that invited respondents to share anything else they could think of
regarding their use (or lack thereof) of RS for planning. The organization that reported
not using RS at all simply shared, “We do not utilize remote sensing.”
One of the municipal planning organizations that demonstrated particularly high
average
RS integration shared that, “Since we have a dedicated GIS employee, the planners do
not have to maintain the city maps or need to manipulate the software, but merely use it
as a tool.” While another highly integrated organization shared that, “Two of our
planners are geographers with technical experience by educational background, so
they have a deep understanding of the value of remote sensing and digital image
analysis.” These responses add context and flavor to some of the integration and
characteristics results.
CHAPTER 5: DISCUSSION AND CONCLUSIONS
Overview
This study focuses on the relationships between integration of RS within an
organizations’ planning practice and their characteristics. It asks multiple research
questions at the level of the metropolitan statistical area, gathering data from municipal
planning organizations across the St.
Louis MSA. Each of these research questions builds on one another in a sequence.
Research Questions:
1. What are the popular RS techniques, technologies, and applications used across
the St. Louis Metropolitan Statistical Area (MSA) for urban planning?
2. Do planners in municipal organizations across the St. Louis MSA believe that
RS is valuable to the urban planning process?
3. What is the current level of RS and planning integration across the St. Louis
MSA, based on the F.A.T.E. Integration Matrix?
4. Do demographic qualities of a municipality or characteristics of the municipal
planning organization have a measurable effect on an organization’s level of RS
integration in urban
planning?
5. Do either organization characteristics or integration scores have a significant
relationship with organization opinions on the integration of RS and urban
planning?
The extensive survey that was created to pursue these five research questions
included sections on the organizations’ preferred RS tools, organization characteristics
(such as financial capabilities, staff makeup, etc.), and planners’ opinions on inclusion
of RS in the urban planning process. The survey also included a set of targeted
questions for measuring each of the four
F.A.T.E. Integration Matrix categories. F.A.T.E. integration scores were then dissected
by employing descriptive statistics and generating a comprehensive average integration
score for each organization. In this way, a four-pronged dataset was generated for each
organization:
1. Popular RS Tools, Applications, and Data Products
2. Municipal Demographics / Organization Characteristics
3. Planners’ Opinions on RS and Urban Planning
4. F.A.T.E. Integration Matrix Scores & Average Integration Scores
To fully address the research questions posed in this study, it required going
beyond simply generating a dataset and describing it. Spearman’s Correlation
Coefficient, selected for its acuity at handling non-parametric bivariate data, was
employed to search for relationships between each element of this quadruple dataset.
The results of the descriptive statistics and the resulting significant bivariate
correlations are provided in the preceding chapter. This discussion chapter will shift to
summarizing key findings, making sense of these findings in the context of previous
literature, and highlighting the study’s limitations and future possibilities.
Survey Responses
Survey responses were submitted by only small- to mid-size cities in the St. Louis
MSA (those under 50,000 population). This presents both a limitation and an
opportunity for this study. As a limitation, being provided with responses from only the
lower population cities means that conclusions are only relevant to that demographic
and not, for instance, the City of St. Louis proper which boasts a municipal population
of over 280,000 (“Census Data,” n.d.). It is also presumed that these larger population
cities may have larger budgets, personnel counts, and thus bigger GIS and planner
departments. Insights about that top range are not capture in this study. However, as
limited literature focuses on the range of city sizes captured in this study, this range of
survey responses provides further opportunity to contribute to the literature.
Techniques, Technologies, and Applications
Strong importance ratings were provided regarding RS value to each of the five
applications provided – Code Enforcement, Drafting the Comprehensive Plan, Creating
Staff Reports, Reviewing Permits, and Creating Maps for Public Use. As no other
applications were provided by respondents, and rankings were provided for these five
across all eleven organizations that use RS, a natural conclusion is that none of the
respondent planners felt the need to indicate any uses for RS beyond those provided.
The universal daily usage of aerial imagery demonstrates its critical role for aiding
the completion of common planning tasks. This finding dovetails nicely with the
universal popularity of County GIS portals, as these often host AIP data for those
counties throughout the St. Louis MSA. AIP’s dominant importance for all five of the
provided planning tasks also suggests that planners find it critical to a broad spectrum
of their workflows. This result also aligns with HoalstPullen and Patterson’s (2011)
finding in the Atlanta, Georgia metropolitan area, with thirteen of fourteen
municipalities utilizing it, and most notably, the only RS platform that municipal
planners in their study reporting using whatsoever. Their result is countered here by
wide municipal use of satellite imagery and some use of drone imagery and LiDAR
presented in the results of this study, though it is important to note that thirteen years
have passed since the publication of their article and the rapid pace of technology
adoption and data sharing could account for the more diverse adoption of RS
technologies found in St. Louis in 2023. Hoalst-Pullen and Patterson (2011) also note
the distribution of AIP data through county-hosted web sources.
The slight lag in usage of satellite imagery (SI), as compared to AIP, could be due
to a few potential contributing factors: 1) SI may not hold the same value for
visualization at the individual property scale, as much of it is ten- or thirty-meter
resolution (as opposed to sub-foot resolution in many cases for AIP), 2) SI is more
likely to be made available on State and Federal web pages, such as GIS data
warehouses or the USGS Earth Explorer application, which both reported very little
usage from respondent planners.
The reported usage of LiDAR appears to be concentrated around two applications
areas: drafting the comprehensive plan and reviewing permits, though it does not
appear to be considered very critical to either. This could suggest a lack of accessibility,
a steep learning curve, or a lack of perceived value for this data product towards the
three other planning applications. LiDAR is a notoriously difficult method to learn,
capture, and analyze, and often requires larger amounts of data storage capacity.
Drone imagery’s moderate performance across all five of the planning applications
could be notable due to DI’s affordability and modularity, though the sample size is too
small to make any strong conclusions for either LiDAR or DI. Drone imagery use
could also be dampened by legal requirements surrounding flying drones, such as the
need to have a UAV pilots license to fly many drones in a professional capacity. DI is
likely the only data product that planners / GIS personnel could capture themselves,
though when viewed through the lenses of associated training and certification cost
associated with generating the dataset, it is not surprising that usage lacks here.
However, the moderate performance across all five planning applications could suggest
reasonable room for growth in its use by planners.
The reported results for the popular tools and data portals used by planning
organizations suggest a trend where planners engage slightly more often with software
capable of viewing RS data than they do software capable of performing data
transformations or analysis on it, possibly due to an overall lack of familiarity with RS
analysis methods. Software / tools for visualizing RS data, such as Google Maps,
demonstrate slightly higher popularity than tools with more analytical capabilities, such
as ESRI ArcMap and ArcGIS Pro or Google Earth Engine. County GIS portals
/ viewers’ universal use among the sample could be related to planners’ familiarity with
this resource due to their common need to reference parcel or land records data, which
is most commonly hosted on county GIS platforms.
In addition, while the write-in options (Nearmap and Eagleview Pictometry) were
not rated on importance to planning tasks, these two options may provide insight into
the needs of planners in this area beyond those options provided in the survey. From an
internet search, it appears that Nearmap advertises itself as provider of on-demand
high-resolution aerial imagery (“Nearmap,” n.d.). Eagleview Pictometry appears to
provided multiple GIS and imagery services, with services for government agencies
advertising high resolution imagery at multiple angles with an associated imagery
viewer / analyst software (“Government,” n.d.). Both write-in results support the idea
that planners value high-resolution aerial imagery and this pairs well with findings
suggesting a slight preference for using RS data to visualize individual properties or
features.
Remote Sensing Value to Planning
Results for organization beliefs statements C (‘My organization would fund training
for planners to learn more about remote sensing’) and G (‘Planners in my organization
believe that remote sensing is a powerful information-gathering tool’) provide evidence
that planners do in fact believe that RS is valuable to urban planning, particularly as an
information-gathering tool that they can use to understand their city’s urban
environment. The presence of this belief among planners is an important finding, as it
acts as a proverbial green light for researchers in the RS and urban planning fields that
there is room for growth and acknowledgment of value. Planners also do show positive
agreement (albeit weakly) to increasing RS use in the future and finding it to be a
valuable financial resource. All these results point towards a fertile landscape for the
growth of RS use for urban planning throughout the study area.
Results for Organization Beliefs Statement E (‘Planners in my organization believe
that remote sensing is convoluted / difficult to use or understand’) also demonstrate
that planners do not find RS to be convoluted or difficult to understand. This result
almost appears to be at odds with the relatively neutral stance of the sample group
towards understanding what the term RS means. These two in combination could
suggest that while planners believe they have a handle on RS as a tool, they may not
identify as having a strong positive understanding of the science or technology under
the umbrella term of RS, though more research is needed here to examine how these
two results could coexist.
This appears to be the first study that sought to measure RS value to planning through
asking planners to agree with Likert-type value statements. The broad nature of the
statements posed in question 15 encourage comparison to other metropolitan areas (or
recreation with other types of organizations).
Measuring RS Integration
The F.A.T.E Integration Matrix was developed by building on the four RS
integration barriers identified in Hoalst-Pullen and Patterson (2011). It provided
empirical findings that are useful for describing the landscape of RS integration by
providing a structure for comparing different levels of integration across the four
dimensions (financial, applications, technical, and expertise). It is important to note that
these four categories may not encompass all the potential barriers to RS integration
with urban planning practice. That is why planners’ opinions and characteristics of their
planning organizations were included in the survey or collected post-hoc. There is
certainly further work to be done to refine not only the F.A.T.E. Integration Matrix, but
also to refine which component Likert statements are used to measure each of the
matrix’s dimensions. While much deliberation went into which Likert-type item
statements were included to measure characteristics of each F.A.T.E. dimension, the
final set of Likert statements is ultimately subjective. For this reason, it was valuable to
include each component Likert statement as its own dataset when searching for
correlations and relationships.
The overall average score for RS indicates promising potential for growth in
integrating RS in planning workflows, with an emphasis on its value in solving
planning problems. This dovetails well with their mild inclination to increase RS use in
the future (see the previous research question discussion). It is important to note that all
eleven organizations that used RS had integrations above zero. Unfortunately, the
Expertise F.A.T.E. Integration Matrix dimension presents a notable challenge based on
a neutral integration score, suggesting a sample-wide lack of expertise as the major
barrier to furthering RS integration in the St. Louis MSA. Financial and technical
integrations show moderate average scores, while applications integration was high.
The strong applications score indicates several applications areas where RS can add
value to planning workflows.
Financial integration, as the second lowest integration score, is highlighted as a
moderate barrier and area of potential growth. This could be due to the high potential
costs associated with RS, including training, data acquisition, and costs associated with
higher computing power needed to perform many versions of RS analysis. Despite this
moderate barrier, the results of this study suggest that organizations in the St. Louis
MSA do not lack financial capability or goodwill towards investing in RS for urban
planning, however, neutral agreement metrics in this dimension suggest it is not a
significant budgeting priority. Responses in the organization beliefs sections suggest
that it may not be a priority in hiring or compensation conversations for planners,
either. The complexity of financial integrations effect on overall integration is not fully
reflected in other work, such as that of Hoalst-Pullen and Patterson (2011), where the
authors concluded that all four of the F.A.T.E. categories except for financial
constraints were a factor in limiting RS adoption. That is a fair assessment, but it fails
to capture the overall lack of financial prioritization of RS training or capabilities, with
financial integration being described as neutral (or “lukewarm”) at best in this study.
Applications integration emerges from the F.A.T.E. Integration Matrix as the
dominant integration dimension. This indicates a strong connection between RS and
common planning tasks and demonstrates that the application potential is known and
valued. Respondents agree that RS adds value to a host of planning workflows,
although creating and drafting plans (especially the comprehensive or ‘master’ plan)
and municipal code enforcement were identified as weaker applications areas compared
to other tasks. This could also suggest that simpler tools may be adequate to complete
these tasks successfully or satisfactorily, which lag in RS performance.
Technical integration, while scoring moderately well, also shows room for
improvement. One potential area of improvement could come from increasing the
willingness of municipal planning organizations to share and collaborate with other
organizations around RS data. This result would be ripe for further exploration by
potentially using interviews and/or focus groups, which could also investigate whether
this is a result specific to RS or present across the sharing of other types of data
between and beyond municipal organizations.
Results for this research question underscore the critical role of expertise and
experience. The expertise integration dimension represents the largest barrier or
potential growth area for furthering integration of RS in planning throughout the
respondents. The metrics for each of the expertise dimension Likert statements
indicated that RS is often excluded in training, onboarding processes, and also
indicated that organizations generally refrain from asking about RS experience in
interviews. There is a clear need for more training and education around RS. There is
also a critical need for municipal organizations that hire planning personnel to ask
about RS experience in the hiring process and inclusion of RS concepts in onboarding
processes if they intend to increase its use. Furthermore, organizations appear willing
to provide funds for technology and training to enable RS use, but there appears to be a
distinct lack of incentives for planners to acquire and nurture these skills.
Relationships: Integration, Organization Characteristics, and Planner
Beliefs
While each F.A.T.E. Integration Matrix Likert statement contributed to their
respective dimensions, many of the most interesting speculations and conclusions came
from the individual correlations between statements in this section, organization
characteristics, and planners’ reported opinions. One notable finding is the significant
negative correlation between organizations’ overall financial integration score and their
total number of GIS personnel employed, meaning that as organization’s overall
financial integration score increased, their total number of GIS personnel decreased.
This result could suggest that municipal organizations with larger GIS departments may
undervalue the financial investment in RS for planners (the focus of this study). This is
possibly due to a perceived redundancy in the skills and technologies already covered
by larger and more capable GIS departments.
The number of GIS personnel also had implications for other aspects of RS
integration, beyond financial components. For example, planners at organizations with
larger GIS and planning departments reported that their organizations were more likely
to fund training for planners in RS. This finding could highlight the importance of
exposure and diversity in educational backgrounds within municipal organizations, as
the educational level of personnel also increased with increases in financial integration
of RS. These results suggest that experienced GIS personnel could be acting as
ambassadors of RS use to leadership and planning staff.
In terms of applications integration, the results indicate that organizations with
more planning personnel tend to explore applications of RS for permit review more
extensively. This could position RS as a tool more commonly used by larger
organizations, maybe with higher bandwidth to explore more analytical applications of
RS. Furthermore, as the percentage of planners in an organization with AICP
certification increased, so did their opinion that RS is beneficial to permit application
review and creating/drafting city plans. This underscores the role of professional
certifications and continuing education in promoting more advanced use of RS for
planning tasks.
The results also revealed that as the size of planning departments increased, so did
their perception of how adequate the computer hardware was with relation to RS. This
result is interesting when contrasted with the finding that as planners’ average years of
education increased, their perception that they had adequate hardware and software
capable of supporting RS analyses decreased. This could suggest a potential mismatch:
planners who are more highly skilled / educated / exposed to RS may be more likely to
characterize their organization’s hardware and software as lacking sufficient
computational capabilities to support more advanced analysis. Perhaps those who have
more experience or knowledge on the topic are more critical of their resources because
they know what they could be missing or what capacities are possible, not to mention
they could have higher base expectations. However, those who have less experience
and knowledge would not know what the options or potential are in the same way.
As GIS personnel education and department size increased, so did planners’
perception of their RS training being sufficient for their level of use. This may indicate
that organizations having larger GIS departments with higher education levels instill
confidence in their planning personnel regarding RS skills. This appears to be
irrespective of planners’ own experience / education level, which did not show a
significant relationship with RS training sufficiency. This finding may underscore the
importance of organization culture and leadership, where GIS personnel reach ‘across
the hall’ to planning department peers to foster a culture of RS integration/proficiency,
and further supports them acting as ambassadors and enablers for RS use by planners.
Overall, the results to Research Question 4 emphasize the complex interplay of
overall organization size, differing education backgrounds, professional certifications,
and financial and technical perceptions in shaping the landscape of RS integration in
urban planning workflows. Larger organizations with more diverse educational
backgrounds and strong GIS department representation tend to lead in all aspects of RS
integration. This highlights a need for investment in RS education, training tailored to
planners, and enhanced technological infrastructure within planning organizations to
better manage RS data. These investments would allow municipal organizations to
fully leverage the potential benefits of RS for urban planning. However, as discussed
previously, the concept of increasing RS in planning logically depends on the amount
of incentives for planners to pursue training and integration of RS in their work. These
incentives can be addressed during the hiring process and as opportunities for training
during and after onboarding emerge through the planner’s tenure at a municipal
planning organization. Moreover, professional organizations and universities can be
critical for encouraging municipal planners to engage with RS as a serious avenue for
broadening their skills as part of formal professional development and maintenance of
credentials such as AICP certification.
When planners reported a stronger understanding of RS it was more likely that
organizations inquired about RS during job interviews. This suggests that prioritizing it
during the recruitment phase could lead to an increased level of knowledge than those
organizations that do not inquire about it when they are hiring. This is further backed
by the finding that organizations who ask about RS in job interviews are more likely to
have planners who recognize the value of
RS for urban planning and who report that there are aims to increase the use of RS in
the future. In other words, whether they valued raster data as highly as vector data or
whether they asked about RS in the hiring process, each had a significant relationship
with whether there were plans to increase RS in the future. This highlights the
importance of making RS use a highly visible and advertised priority and suggests
another area that professional organizations (such as the APA) could encourage
planners to explore further in their studies.
Increases in average overall integration were related to both a decrease in thinking RS
was convoluted / difficult to understand and an increase in the perception that RS is a
powerful information gathering tool. This suggests a barrier for entry due to the
perceived complexity of RS concepts, despite RS being widely recognized for value to
municipal planning processes.
Planners may advocate for RS when they believe it is valuable and a financially
worthwhile investment. Planners with more training and education with RS are more
likely to also hold the perception that their hardware and software is less adequate to
complete RS tasks. The lesssufficient technical capabilities reported within highly
educated planning departments could be caused by their identification of more (or more
advanced) planning applications for RS that require additional software or increased
hardware capabilities. This underscores the importance of planners advocating for
sufficient resources to leverage RS effectively.
These results underscore the importance of RS not just as valuable data for planning
processes, but as a tool that planners use to empower themselves in their role. This
belief’s positive correlations with technical capabilities and overall RS integration
scores highlight the significance of empowering planners to complete RS tasks through
increasing their capacity to manage RS data and setting them up for success. Also,
when planners have perceived financial support to attend RS trainings, the significant
negative relationship to the perception that RS should be outsourced suggests that
access to training could provide planners with the skills and confidence to support RS
tasks in-house.
Limitations and Methodological Reflections
One assumption of this study is that planners and their organizations would benefit
from increasing RS use in the future. Many of the recommendations stemming from the
results of this study surround enabling planners and their municipalities to capitalize on
further investment in RS. However, it is important to note that one of the twelve
respondents indicated that they are not using RS, that they know no value for its use for
urban planning, and that they do not plan to increase its use in the future. The
assumption that RS use is generally productive for planners is made across the
literature. One of the driving forces behind the pursuit of this set of research questions
is a general lack of understanding of how planners feel about RS use in urban planning
and whether they can actualize any perceived value in its integration with their process.
Increasing its use, reducing barriers for further integration, and finding new
applications for RS in urban planning may not be a fit (or a desire) for every planning
organization, and it is important to acknowledge that up front.
This thesis employed a comprehensive online survey as the primary instrument.
One limitation of the survey was that it did not provide the freedom to follow up and
explore in more depth any of the interesting findings in the data. The lack of follow-up
potential presents an opportunity to build on this study using interviews and/or focus
groups to dig deeper into the interesting patterns uncovered by this initial study.
Furthermore, the qualitative data analysis portion of this study was severely limited by
the lack of respondents’ answers in the open-ended write-in portions of the survey. Had
more participants chosen to add answers to those parts of the survey there would have
been a greater amount of qualitative material to code and analyze. Coding could have
been used to support interpretations of the data and conclusions. For these reasons,
future studies may benefit from adding a place in the survey where participants could
opt to provide their contact information and volunteer to participate in follow-up
interviews and/or focus groups that could leverage and explore emerging trends and
patterns found in this study.
Multiple research questions were designed to search for correlations between
different sets of data, including planner’s opinions, organization characteristics, and
F.A.T.E. Integration Matrix characteristics. While there were many strong correlations
and the directionality of these correlations could be determined using the Spearman’s
Correlation Coefficient, it is important to remember that correlation does not indicate
causation. The discussion portion of this thesis includes many speculations at
causation, but these are just that. This provides another area for building on this study:
leveraging the strong relationships indicated in the results of this study and following
up on them with more qualitative methods to explore causality. These could include
follow-up phone or in-person interviews to capture planner’s nuanced reflections on
results from the survey.
Many of the follow-up questions which organically arose from discussing the
results of the survey relate to digging deeper into planner’s educational backgrounds.
There appears to be a disparity between RS integration between those planners with
GIS / RS experience and training and those without. While the survey implemented in
this study asked about planner’s years of education and their highest degree achieved, it
did not ask about degree type or year attained. It would be interesting to study whether
planners with university geography or GIS degrees tend to implement RS, or whether
planners with more recent degree attainment have more experience and understanding
of RS.
The length of the survey, while field tested and found to be completable in
approximately fifteen minutes, could have contributed to the low response rate (n=12;
13.8%), supported by the fact that there was evidence of starting and not completing
the survey. Ensuring that there is a visible progression bar on the survey in the future
may also encourage more people who start the survey to finish. Moreover, the study’s
focus on RS (commonly perceived as a technical science or specialty) may have also
contributed to the low response rate. While significant correlations were found,
marketing the survey to a broader audience may have benefited this study by providing
a larger pool from which to sample, thereby opening the door for additional analyses.
This study was geographically delimited to the St. Louis MSA, to replicate and
build on a previous study focused on the metropolitan area surrounding Atlanta,
Georgia (Hoalst-Pullen and Patterson 2011). While Hoalst-Pullen and Patterson (2011)
were able to gather data from a wide range of organization sizes and types (municipal,
county, regional, etc.), this study only represented responses from mid or small sized
municipalities (those of 50,000 population or less). Many highly populated
communities exist in the St. Louis MSA, but these did not submit responses for
unknown reasons.
Solicitations to partake in future research should also explore leveraging
professional planning organizations, such as this study did with the St. Louis Metro
American Planning Association Chapter. Assisted solicitation by this professional
organization nearly doubled the small sample size, and this would likely only improve
at the state or national organization level. Asking administrative staff at the
organizations for their help in identifying the appropriate person to send the survey to
may also boost response rates.
Conclusions
This study explored the integration of RS with urban planning in municipal
organizations across the St. Louis MSA. It focused on research questions related to
popular RS techniques, technologies, applications. A comprehensive survey was
deployed which collected data on organization characteristics, popular RS tools,
integration levels, and planners’ opinions towards RS for urban planning applications.
This study contributes to a deeper understanding of the adoption of RS by actual
practitioners in the urban planning field.
Remote Sensing Integration with Urban Planning
This study formalized the integration barrier framework presented by Hoalst-
Pullen and Patterson in 2011 and developed it into the F.A.T.E. Integration Matrix,
which reveals varying levels of integration across financial, applications, technical, and
expertise RS integration dimensions among municipal planning organizations in the St.
Louis MSA. Exceedingly high applications integration scores across the entire dataset
demonstrate the strong need that planners feel for RS to help them with fulfilling their
professional roles, however, the other three dimensions showed lacking integration,
leading to the following key findings for integration improvement:
•Expertise integration scored lower than the other three F.A.T.E. dimensions.
Planners (and the leadership of organizations that support them) should invest
in more professional development or continuing education and training around
RS use. There is also a critical lack of RS acknowledgement in processes for
interviewing prospective planner job candidates and in organizations’
onboarding process for new planner hires.
•Technical integration can be improved across the region. This integration
dimension, as measured here, was significantly weakened by a lack of
collaboration around RS data by planning organizations, as well as a moderate
lack of computational power available to planners for RS tasks.
•To improve financial integration of RS in planning, planners and their
organizations should make financial investment in RS technologies, trainings,
and tools a priority in budgeting discussions and provide financial incentives
for planners to increase their RS knowledge such as increased compensation
and training opportunities.
Municipal / Organization Characteristics Influences
This study also explored the impact of organization characteristics variables such
as geographic size, personnel education levels, and GIS and planning department size
on RS integration scores. Some of these variables exhibited strong correlations with RS
integration levels, indicating that organization characteristics do indeed play a role in
shaping the adoption of RS with urban planning at municipal organizations in the St.
Louis MSA.
The negative correlation between GIS department size and an organization’s
financial integration suggests that larger GIS departments may reduce the need to
harbor RS skills within organizations’ planning departments. Larger departments with
more planning personnel tend to explore RS applications for permit review more
extensively, especially if these planning personnel pursue continuing education such as
the AICP certification. Technical integration findings demonstrate a positive correlation
between planning department size and planners’ ability to procure the hardware they
need to complete RS tasks, yet the opposite is true for more experienced planning
departments, possibly due to their desire to pursue more advanced applications of RS.
Expertise integration reveals that organizations with experienced GIS departments may
instill confidence in planning personnel, potentially emphasizing the relationship
between organization culture and technology adoption.
Overall, these findings underscore the need for continued investment in municipal
GIS departments, and for planners to be given access to tailored RS training and
enhanced technological infrastructure to fully take advantage of the benefits of RS for
planning workflows.
Planners Opinions
Through Likert-type scale analysis, this study gained insights into planners'
opinions regarding RS’s value, challenges, and prospects in urban planning.
Understanding these perspectives is crucial for aligning technology adoption with
practitioners' needs and expectations. Results showed that planners do indeed value RS
for urban planning, especially as an informationgathering tool. They do not find it
convoluted, though they may struggle to understand what all the term RS encompasses.
Popular Technologies
Four main RS data products were analyzed, with AIP emerging as universally used
daily, indicating its potentially critical role in regular planning tasks among
respondents. SI followed closely, with strong, frequent use, but lagging AIP, potentially
due to its lower resolution for individual property-scale visualization. LiDAR use
showed potential for drafting comprehensive plans and reviewing permits but saw
significantly lower and less-frequent use that SI. DI, while the least commonly used
among the four data products, demonstrated moderate performance across all the
various planning applications areas, suggesting some room for growth in its future
utilization.
The survey also highlighted a moderate preference among planners for software
platforms capable of viewing RS data over tools more equipped for RS data analysis.
County GIS
Portals/Viewers were universally used, indicating familiarity and ease of access among
planners. Google Maps and similar visualization tools/software were more popular than
analytical RS tool suites such as ESRI ArcMap or Google Earth Engine. The discussion
points to the need for further investigation into why some RS products are favored over
others. Additionally, two write-in software options (Nearmap and Eagleview
Pictometry) suggest a demand for high-resolution aerial imagery access beyond the
provided options. While the survey asks about the relative value and frequency behind
these patterns of usage, it did not dig into why these patterns exist.
Implications and Recommendations
Based on these findings, several implications and recommendations emerge:
•Capacity Building: Organizations with lower integration scores may benefit
from targeted capacity-building efforts. These can include incorporating
questions about RS experience in the interview process, adding RS components
to onboarding and training programs, increasing resource allocation for RS
tools, and incentivizing staff to grow and maintain RS skillsets.
•The planning community should leverage the support of universities and
professional organizations such as the APA to establish formalized training and
experience pathways
for planners seeking to increase their understanding and implementation of RS
technology in their execution of their roles.
•Collaborative Initiatives: Encourage collaboration among planning
organizations, industry experts, and academic institutions. This can foster
greater knowledge exchange, the sharing of best practices, and collective
problem-solving through integrating RS technology and techniques with
planners’ needs and workflows effectively.
•Data Standardization: Standardizing data collection, analysis methodologies,
and reporting mechanisms for measuring integration of RS in planning can
enhance comparability of results between academic studies and set benchmarks
for integration efforts across different organizations. This would facilitate the
tracking of further integration of RS, and the progress of RS adoption within
municipalities across metropolitan study areas.
•Continuous Evaluation: The F.A.T.E. Integration Matrix presented in this study
is a pilot framework, one which builds on previous work by Hoalst-Pullen and
Patterson (2011) through establishing Likert-type item matrices to measure the
integration of RS in four key areas: financial, applications, technical, and
expertise integration. The Likert-type item matrices can benefit from continued
scrutiny and evaluation which will help fine-tune the future use of this kind of
analysis for measuring organizations’ integration of RS in urban planning.
Future Research
This study captured data from a dozen municipalities within the St. Louis MSA.
The methods employed in this study for measuring RS integration, while based on
dimensions of RS integration proposed by Hoalst-Pullen and Patterson in 2011, are
novel, easily modifiable, and reproducible.
This study design opens the doors to comparing different study areas, planning
organization types, and region sizes. Future research could expand the scope to include
a larger sample size, diverse geographical regions, longitudinal studies for tracking
integration trends over time, and qualitative methods for deeper insights into
stakeholders' experiences and perspectives. Modifications to the survey format are also
proposed in the discussion, including adding more open-ended questions and
encouraging more qualitative data production from respondents to dig even deeper into
underlying motivations and limitations surrounding RS adoption by working planners.
Conclusion Statement
In conclusion, this study sheds light on the current landscape of RS integration in
urban planning organizations in the St. Louis MSA. By measuring integration
characteristics across multiple municipal planning organizations, leveraging insights on
organization characteristics, and aligning with planners' needs, we can pave the way for
enhanced utilization of RS technology to drive more informed, sustainable, and
effective urban planning practices. This study should be one of many future endeavors
exploring the needs, desires, attitudes, barriers, and adoption patterns of planners as it
pertains to integrating RS with their daily work. By aligning these, more important and
innovative academic research can reach its target audience, and RS can further integrate
with planning practices, creating more informed municipal decision-making bodies
across the world, and ultimately benefiting residents.
Overall, the results of this study suggest a promising landscape for RS integration
with municipal urban planning. The results also highlight key areas that these
organizations can target to improve this integration. Municipal planning organizations
in the St. Louis MSA should focus on addressing existing RS expertise gaps, improving
their organizations’ technical capabilities (where needed), sharing RS data / analysis,
and leveraging the clear value of applying RS to addressing planning challenges and
within planning workflows. Organizations which aim to increase the use of RS should
consider discussing RS in their training and onboarding processes, as well as seeking
candidates with RS experience when interviewing candidates for planning positions.
