Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Developing a replicable approach for the creation of urban climatic maps for urban heat island analysis: a case study for the city of Los Angeles, California
(USC Thesis Other)
Developing a replicable approach for the creation of urban climatic maps for urban heat island analysis: a case study for the city of Los Angeles, California
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Developing a Replicable Approach for the Creation of Urban Climatic Maps for Urban Heat
Island Analysis: A Case Study for the City of Los Angeles, California
by
Bryan Gene Lam
A Thesis Presented to the
FACULTY OF THE USC DORNSIFE COLLEGE OF LETTERS, ARTS AND SCIENCES
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(GEOGRAPHIC INFORMATION SCIENCE AND TECHNOLOGY)
August 2020
Copyright © 2020 Bryan Gene Lam
ii
Acknowledgements
I would like to thank Dr. Andrew Marx for his help and guidance into making this project
possible, and to Dr. Darren Ruddell and Dr. Steven Fleming for their expertise and comments
into shaping and improving this project. I also appreciate the advice Dr. Robert Vos gave me to
help steer my project into the right direction. I would also like to thank my coworkers and
supervisors at the Los Angeles Department of City Planning and Department of General Services
for technical help and for allowing me to use the data necessary for this project. I extend special
thanks to Caroline Seymour for her expertise on the urban heat island and to Juan Chavez for his
encouragement to pursue this topic. Lastly, I would like to thank my parents for their continued
support all these years and am grateful for everything they have done for me.
iii
Table of Contents
Acknowledgements .................................................................................................................... ii
List of Tables ..............................................................................................................................v
List of Figures ........................................................................................................................... vi
List of Abbreviations............................................................................................................... viii
Abstract .................................................................................................................................... ix
Chapter 1 - Introduction ..............................................................................................................1
1.1. Motivations .................................................................................................................... 1
1.2. Research Objectives ....................................................................................................... 5
1.3. Study Area...................................................................................................................... 6
1.4. Thesis Organization ........................................................................................................ 9
Chapter 2 - Related Work .......................................................................................................... 10
2.1. Urban Heat Islands ....................................................................................................... 10
2.2. Urban Heat Island Studies in Los Angeles .................................................................... 13
2.3. Standardizing Urban Heat Island Studies ...................................................................... 15
2.4. Urban Climatic Map ..................................................................................................... 16
2.5. Urban Heat Island Mitigation Strategies ....................................................................... 19
Chapter 3 – Methods ................................................................................................................. 23
3.1. Data Sources................................................................................................................. 23
3.1.1. Los Angeles Region Imagery Acquisition Consortium Datasets ........................... 24
3.1.2. Government Datasets ........................................................................................... 25
3.2. Research Design ........................................................................................................... 27
3.2.1. Thermal Load Map: ............................................................................................. 29
3.2.2. Dynamic Potential Map:....................................................................................... 35
3.2.3. Wind Data............................................................................................................ 43
iv
3.2.4. Urban Climatic Map ............................................................................................ 46
3.2.5. Validation and Linear Regression......................................................................... 50
Chapter 4 - Results .................................................................................................................... 51
4.1. Thermal Load Map ....................................................................................................... 51
4.2. Dynamic Potential Map ................................................................................................ 56
4.3. Urban Climatic Map without the Wind Layer ............................................................... 64
4.4. Validation/Regression Map........................................................................................... 66
4.5. Urban Climatic Map with Wind Layer and Prevailing Wind Information ...................... 70
Chapter 5 - Discussion and Conclusion ..................................................................................... 78
5.1. Observations ................................................................................................................. 78
5.2. Limitations ................................................................................................................... 81
5.3. Future Work ................................................................................................................. 87
5.4. Conclusion ................................................................................................................... 92
References ................................................................................................................................ 93
5.5. Bibliography ................................................................................................................. 97
v
List of Tables
Table 1: Summary of Government Sourced Datasets ................................................................. 26
Table 2: Building Volume Thermal Load Classification (Taken from Ng and Ren 2012) .......... 31
Table 3: Elevation Thermal Load Classification (Taken from Ng and Ren 2012) ....................... 33
Table 4: Vegetation Thermal Load Classification (Taken from Ng and Ren 2012) .................... 35
Table 5: Ground Coverage Dynamic Potential Classification (From Ng and Ren 2012) ............. 36
Table 6: Green Space Dynamic Potential Classification (Taken from Ng and Ren 2012) ........... 38
Table 7: Distance from Coastline Dynamic Potential Classification (From Ng and Ren 2012) ... 40
Table 8: Open Space Dynamic Potential Classification (Taken from Ng and Ren 2012) ............ 41
Table 9: Slope Dynamic Potential Classification (Taken from Ng and Ren 2012) ...................... 42
Table 10: NOAA NCEI Weather Stations ................................................................................. 44
Table 11: Urban Climate Variable Summary ............................................................................. 47
Table 12: Urban Climate Classification Descriptions ................................................................. 49
vi
List of Figures
Figure 1: City of Los Angeles .....................................................................................................7
Figure 2: Los Angeles urban climatic map methodology ........................................................... 28
Figure 3: Building Volume Layer Methodology ........................................................................ 31
Figure 4: Elevation Layer Methodology .................................................................................... 33
Figure 5: Vegetation Layer Methodology .................................................................................. 34
Figure 6: Ground Coverage Layer Methodology ....................................................................... 36
Figure 7: Natural Landscape Layer Methodology ...................................................................... 38
Figure 8: Proximity to Waterbodies Layer Methodology ........................................................... 39
Figure 9: Proximity to Open Spaces Layer Methodology ........................................................... 41
Figure 10: Proximity to Open Spaces Layer Methodology ......................................................... 42
Figure 11: Example Wind Rose Python script ........................................................................... 45
Figure 12: Summer 2017 USC Wind Rose in miles per hour ..................................................... 46
Figure 13: Building Volume ...................................................................................................... 52
Figure 14: Topography .............................................................................................................. 53
Figure 15: Vegetation ................................................................................................................ 54
Figure 16: Thermal Load Map ................................................................................................... 55
Figure 17: Ground Coverage ..................................................................................................... 57
Figure 18: Natural Landscape Cover ......................................................................................... 58
Figure 19: Proximity to Waterfront ........................................................................................... 59
Figure 20: Proximity to Open Spaces ........................................................................................ 60
Figure 21: Vegetated Slopes ...................................................................................................... 61
Figure 22: Proximity to Openness ............................................................................................. 62
vii
Figure 23: Dynamic Potential Map ............................................................................................ 63
Figure 24: Draft Urban Climatic Map ........................................................................................ 65
Figure 25: Ordinary Least Squares result for CalEPA UHII and Aggregated Urban Climatic
Classes ...................................................................................................................................... 66
Figure 26: Ordinary Least Squares result for CalEPA UHII, Aggregated Urban Climatic
Classes, and Projected Wind Speed Values ............................................................................... 67
Figure 28: Summer Months Wind Roses ................................................................................... 71
Figure 29: Santa Ana Wind Months Wind Roses ....................................................................... 72
Figure 30: Weather Station Locations ........................................................................................ 73
Figure 31: Summer Month Wind Roses ..................................................................................... 74
Figure 32: Santa Ana Months Wind Roses ................................................................................ 75
Figure 33: Final Urban Climatic Map without Wind Layer ........................................................ 76
Figure 34: Final Urban Climatic Map with Wind Layer ............................................................. 77
viii
List of Abbreviations
CalEPA California Environmental Protection Agency
ECOSTRESS Ecosystem Spaceborne Thermal Radiometer Experiment on Space Station
EPA Environmental Protection Agency
GIS Geographic Information System
ISA Impervious Surface Area
LARIAC Los Angeles Region Imagery Acquisition Consortium
LCD Local Climatological Dataset
LCZ Local Climate Zone
LST Land Surface Temperature
NASA National Aeronautics and Space Administration
NCEI National Centers for Environmental Information
NOAA National Oceanic and Atmospheric Administration
NREL National Renewable Energy Laboratory
TOC Transit Oriented Communities
UC/UCM Urban Climate/Urban Climatic Map
UHI Urban Heat Island Effect
UHII Urban Heat Island Index
ix
Abstract
Urbanization and other anthropogenic developments have changed the environments we
live in. One of the effects of urbanization is the Urban Heat Island effect, a phenomenon where
urbanized areas experience higher temperatures than surrounding rural, less developed areas.
While an urban climatic map can be a useful tool for understanding the Urban Heat Island effect,
there is not consistent methodology for its creation and implementation in urban planning. This
reduces its utility in informing policy and decision making in Urban Heat Island mitigation
efforts. In response, this study demonstrates an approach for the creation of accurate urban
climatic maps, which can be replicated for all of California. Using the city of Los Angeles as an
example, this approach successfully produced different urban climates, and estimates how each
urban climate affects the Urban Heat Island effect. The urban climatic map overlays various
classified layers from multiple fields, such as urban planning and climatology, to construct the
climatic classes for the city. These classification values are based off of the values used in the
Hong Kong urban climatic map. Each climatic class is a description of the thermal load and
dynamic (air movement) potential that is experienced in each area of the city. Classes with high
thermal loads and low dynamic potentials can be identified as areas experiencing a more intense
urban heat island effect. Using the same methodology, the creation for additional urban climatic
maps, or regional climatic maps is possible, greatly improving regional efforts to mitigate Urban
Heat Island effects across California.
1
Chapter 1 - Introduction
The urban heat island (UHI) effect is a phenomenon in which urbanized areas experience
warmer temperatures than surrounding areas due to urbanization and other anthropogenic
activities. Warmer temperatures influence the living environments and conditions of more than
half of the world’s population living in urbanized and developed areas. Researchers have studied
the urban heat island extensively and have developed various methods to analyze the causes and
effects of the phenomenon. A method used to study the urban heat island effect is using an
urban climatic map. Developed by German researchers in the late 20
th
century, the urban climatic
map is a useful tool for city planners to visualize the urban morphology of their city and the built
environment’s potential to contribute to the urban heat island effect. The urban climatic map can
be further analyzed by urban planners to recommend best use practices and create policies to
mitigate the effects of the urban heat island effect. Population growth, urbanization, and other
developments in major cities, such as Los Angeles, are expected to intensify the effects of the
urban heat island in the future. The purpose of the urban climatic map is to visualize and analyze
the urban heat island effect for Los Angeles, California.
1.1. Motivations
Analyzing the urban heat island effect is crucial in order to understand how urbanization
and development can impact its surrounding environment. As a result of increasing urbanization,
humans have altered the environment and landscape in a way which affects the microclimates of
a region. These effects on the microclimate include altering the region’s air and surface
temperatures, air pollution levels, and the population’s health. It is widely accepted knowledge
that the urban heat island effect contributes to heat and pollution retention within a city, which
2
can have negative implications on the health of the population (Dousset 1992). These effects are
intensified through population growth, land use, and other anthropogenic changes within a city.
Understanding the causes and effects of the urban heat island is crucial to mitigating the negative
warming effects caused by urban development. It is especially important to analyze the urban
heat island in major cities, such as Los Angeles, due to its unique geography and past
developments which contribute to the effect.
The geography of Los Angeles is a major influence in the temperature variations
throughout the city, which can affect the intensities of urban heat island in different regions of
the city. Areas along the coast, such as the Westside experience lower temperatures as a result of
the constant sea-breeze which blows through the region and cools the region. Inland areas within
the Los Angeles Basin, such as South Los Angeles experience much higher temperatures
compared to the rest of the city. Other areas, like the San Fernando Valley, typically experience
higher temperatures due to the topography, where the Santa Monica Mountains prevents the sea
breeze from blowing into the San Fernando Valley. The surrounding mountain ranges effectively
forms a shield, preventing cooler air from the ocean to move into the valley.
The topography of Los Angeles also affects the distribution of development within the
city, with most of the city’s commercial and industrial development found in the flat areas of the
Los Angeles Basin. Residential development, on the other hand, are spread out across the various
terrain of Los Angeles. The typical commercial and industrial development in the city tend to
have a much higher warming potential than residential areas, as there are fewer trees and natural
land cover to provide cool areas. Residential areas, especially in hillside communities in the San
Fernando Valley, typically contains more trees and more natural land cover, which in turn
increases the cooling potential in residential areas. The intensities of urban heat island can differ
3
based on the geography of the area, and these effects can also be influenced by the different land
uses found in the city.
Commercial and industrial development in the city have contributed to making Los
Angeles one of the largest cities in the United States, with a majority of the city covered in
impervious surfaces, which intensifies the urban heat island effect in these urbanized areas.
Anthropogenic materials, such as asphalt and concrete, absorb solar radiation at a much greater
rate and quantity than natural land cover such as trees, grasses, and shrubs. The difference in
surface energy balance has been well documented in many studies and UHI has been recognized
to have an impact in heavily urbanized areas (Arnfield 2003; Taha 1997; Oke 1998).
Concentrations of higher density development in regions such as the Downtown area, Wilshire
Corridor, contribute to the urban heat island by absorbing solar radiation and retaining the
warmer air within the area. Tall buildings, dense urban development, and the limited number of
natural spaces in the city also restrict air flow, reducing the ability for cooler air to ventilate the
area.
An example of a recent development intensifying the urban heat island effect is the
Transit Oriented Communities (TOC) program, intended to provide dense, and affordable
housing around transit lines. TOCs are classified based on tiers and are determined based on a
location’s proximity to a major transit stop and the type of major transit stop. Each tier has
different program guidelines, regulations, and incentives to develop affordable housing within
these areas (LAMC 12.22 A.31 Sec 6). The taller buildings built under this program have a
higher potential to retain heat and restrict air flow to move the warmer air away from the area,
intensifying the urban heat island. It is important to study the warming effects caused by urban
4
development so mitigation strategies can be implemented during the developments’ planning
phase to lessen the effects of the urban heat island.
Los Angeles has been the study site for many urban heat island studies, and these effects
on the city has been well documented. The state of California’s Environmental Protection
Agency (CalEPA) completed a report which visualized the urban heat island effect in many
Californian cities, including Los Angeles. The CalEPA report’s methodology followed many
other UHI studies, in which the exchanges in water and energy between the atmosphere and the
ground and temperature anomalies in an urban setting were analyzed to quantify the UHI effect
(Arnfield 2003). Other studies, such as Tran et al. (2006), tie the urban heat island effect to urban
development, economic, and industrial growth, based on the increase in energy and material
consumption. Climatic mapping is another approach in visualizing different microclimates and
the urban heat island. These climatic mapping tools can guide planners to better create and
develop policies and recommendations to mitigate these effects in the city. Ultimately, the urban
climatic map utilizes datasets that are familiar to urban planners, such as land use and building
form, which can allow planners better understanding the impacts of the urban heat island.
Urban climatic maps have been developed in other cities around the world including in
Europe and Asia. Urban climatic maps were first created in German cities such as Berlin,
München (Munich), and Stuttgart in the period between 1970 – 1990. More recent studies have
utilized this tool to visualize the urban climate for cities in Asia, such as Tokyo, Kaohsiung, and
Hong Kong (Baumüller et al. 1992) (Ren et al. 2013; Ng et al. 2009). Despite its long history
and utilization by many cities, an urban climatic map has not been developed for Los Angeles to
quantify the effects of UHI in the city.
5
1.2. Research Objectives
The main purpose of the Los Angeles urban climatic map is to quantify the urban heat
island effect for the city of Los Angeles, creating an approach that is repeatable assuming
required data is available. Analyses includes identifying specific areas of Los Angeles
experiencing much more intense urban heat island effect than other areas. One of the possible
outputs of the urban climatic map is the quantification and classification of thermal stress and
potential for air movement, as shown in the urban climatic map developed by Ng and Ren (2012)
for Hong Kong. The results of the urban climatic map can aid urban planners and policy makers
in identifying where the most intense urban heat island effects are occurring.
Another objective of the urban climatic map for Los Angeles is to develop a methodology
that can be replicated for other regions in Southern California to analyze their urban climates and
the potentials for urban heat island. Temperature based urban heat island studies often are
specific to a given region due to the variations in the environment. The urban climatic map can
be used as the tool to standardize future urban heat island studies in the future, where it takes an
urban climate classification approach, rather than a temperature-based approach. Future studies
can utilize an urban climatic map to be able to visualize the effects of UHI in proposed future
developments, such as those in transit-oriented communities. With this information, urban
planners and other policy makers can analyze specific causes and effects of the urban heat island
effect at the neighborhood level and propose solutions to mitigate these effects in these specific
areas.
While urban climatic maps have been successfully created for multiple cities around the
world using similar datasets and variables, replicating one specific model for multiple cities
would lead to inaccurate urban classifications. Differences in environment, geographies, and
6
built environment makes it necessary to use datasets and variables which may be unique to a
specific region. As with most cities, the many variables that influence the urban climates and
associated microclimates makes modeling these aspects a complex task. It is important to
understand that one methodology cannot accurately describe the urban climate of multiple cities
around the world. To correct this, the final objective of the urban climatic map for Los Angeles
is to identify possible limitations and inaccuracies, and to propose future work to accurately
model the urban climate of Los Angeles.
1.3. Study Area
The city of Los Angeles is located in Los Angeles County and serves as the county seat.
It is centered around coordinates 34.019394 N, 118.410825 W. The city has a population of
approximately 3,990,456 and has an area of approximately 469 square miles/1214 square
kilometers (U.S. Census, 2018). The city of Los Angeles is bounded by the Santa Susana
Mountains and the San Gabriel Mountains in the north and by the Pacific Ocean along the
western and southern portions of the city. A thin corridor connects Los Angeles with the
southern coastal areas of San Pedro and Wilmington. The rest of the city is bordered by other
cities and unincorporated county areas, which gives the city a unique shape. Figure 1 shows the
boundaries of the City of Los Angeles.
7
Figure 1: City of Los Angeles
8
The city’s diverse geography allows different types of development found within the
city. The mountain, coastal, and basin areas of Los Angeles contains distinct urban development
styles. The Los Angeles basin is relatively flat and is where most of the commercial and
industrial development in the city is located. The mountain areas within the city of Los Angeles
are often preserved and maintained as open spaces or used to develop high-income residential
neighborhoods. The terrain makes it difficult to build dense commercial and industrial
development typically found in the Los Angeles basin area. The geographic variations found in
the city often has various effects on the urban heat island effect observed in the city.
The diversity of the environment can affect the climate of the city. According to the
Köppen-Geiger climate classification system, Los Angeles has a Mediterranean climate,
characterized by hot, dry summers and mild winter seasons (Peel et al. 2007). The seasonal Santa
Ana winds also affect the temperatures and humidity during the autumn and early winter months
in the city. Santa Ana winds bring extremely dry winds to the Los Angeles region, and often
cause wildfires in the mountains surrounding Los Angeles (Westerling et al. 2011).
Microclimates are also found within the city itself, where temperature ranges can vary from
cooler temperatures along the coast to much warmer temperatures found within the San
Fernando Valley and in basin neighborhoods such as South Los Angeles.
9
1.4. Thesis Organization
The proposed thesis is organized into five chapters. Chapter Two reviews a selection of
urban heat island studies conducted over many years to provide a background of the urban heat
island, the causes of the effect, and the various methodologies that have been proposed to
accurately assess the urban heat island effect. This chapter also describes the urban climatic map,
its history, and its significance to urban heat island studies. Chapter Three describes the
methodology for construction the Los Angeles urban climatic map and the data sources required.
Chapter Four presents the results of the urban climatic map and data comparisons to other urban
heat island studies. Chapter Five analyzes and discusses the results of the urban climatic map,
and their contribution to existing research on urban land cover change over time. It will also
contain various limitations of the thesis and presents future work to improve the urban climatic
map for Los Angeles.
10
Chapter 2 - Related Work
The related works chapter reviews various literature to understand the UHI, its
significance, and how the urban climatic map can serve as a method for UHI analysis. The
chapter is divided into five main sections: (1) A description of the UHI, (2) An analysis of the
UHI, (3) The standardization of UHI, (4) The urban climatic map, and (5) Strategies to mitigate
UHI. The chapter is aimed to give context and reasoning for the use of the urban climatic map
for UHI studies in Los Angeles, California.
2.1. Urban Heat Islands
The urban heat island (UHI) is an urbanized or developed area that experiences higher
temperatures than its surroundings. It is the result of anthropogenic changes of the Earth’s
surface and atmosphere, which allows for the observable temperature differences between urban
and rural areas (Oke 1995). The UHI is one of the most studied phenomena and its significance
to various Earth System processes, such as climate change, has been well documented.
According to Parker (2009), the UHI is one of the main focal points that must be addressed in
studies on climate change mitigation strategies. Studies have also analyzed how the UHI can
affect the populations living in urban centers, where the increase in temperatures can affect the
population’s levels of thermal comfort and health.
One of the first studies mentioning the UHI effect was in a study on snowfall frequency
in England, where observations of the effects of artificial warming in London and how it may
affect the snowfall in the city (Manley 1953). Manley (1953) was one of the first studies to
propose the idea of an artificial warming, and the effects on urbanized areas. The study found
11
that the artificial warming effect may affect the snow frequency observed within the inner
London area, but further studies would be needed to support this claim.
In later studies, temperature increases as a result of the UHI can alter climatic patterns in
the atmosphere. These climatic patterns can describe conditions such temperature, wind
conditions, and precipitation. In a study by Oke (1995), the urban airflow and air pollution found
in an urbanized area can influence the ability for clouds to form over a city. As the air heats up
due to the UHI, it has a higher potential to contain more water content. When the warmer air
rises in the atmosphere, it cools, and the water vapor condenses to produce clouds. Van
Heerwaarden and Vilà-Guerau de Arellano (2008) analyzed the effects of urbanization and the
UHI on the humidity and the increased ability for heterogeneous surfaces to form clouds. The
climatic impacts of UHI as a result of increasing temperatures in cities continue to be a
motivation for future UHI studies.
Calculating temperature differences between rural and urban areas are one of the most
common methods to quantify the effects of UHI. The calculated temperature differences between
urbanized areas and less developed rural areas show that urbanized areas experience warmer
temperatures than rural areas due to urban development. These modifications of urban surface
causes surrounding air and surface temperatures to rise several degrees higher than the
simultaneous temperatures of the surrounding rural areas (Tran et al. 2006). Landsberg (1981)
also summarizes various UHI studies of cities around the world and finds that temperatures have
increased as a result of the growing population and city sizes. The temperature measurements for
UHI studies can be measured either through the minimum and maximum temperature differences
in an area, or derived from other datasets, such as from specific land cover such as vegetation.
12
Vegetative cover in a location can affect the temperature of the surrounding environment,
as vegetation has a cooling effect on the surrounding environment. The Normalized Difference
Vegetation Index (NDVI), a remote sensing index, is one of the most commonly used remote
sensing index to quantify the amount of vegetation detected in an area. Since vegetation has the
potential to mitigate UHI effects, researchers can use the NDVI to measure temperature changes
based on vegetation density. Weng et al. (2004) takes a different approach and utilizes a form of
vegetation fraction index that is based on land cover type to measure land surface temperatures at
the city block level. As a result of urbanization, vegetative land cover is often removed to
provide space to build infrastructure. This leads to decreases in effects such as evapotranspiration
and solar irradiance, which influence the temperatures in the surrounding area.
Another method of calculating the effects of UHI is the use of surface energy balance
models. The variability in how an area can absorb, reflect, and retain solar radiation is the main
driver of the UHI effect in a given area. According to Arnfield (2013), soil albedo, moisture
availability, and other aspects of the water cycle in the surrounding environment are important
components in the surface energy balance and its role in the warming of a region. This would
include an increase in solar radiation that is being absorbed by anthropogenic development and
the reduction of evapotranspiration rates between the land surface and the atmosphere, which
increases temperatures (Kim et al. 1992). Altering the environment with anthropogenic
infrastructure such as buildings, roads, and other anthropogenic development can lead to solar
radiation at a rate much higher than the rate found in rural, open areas.
In other studies, the surrounding biomes and environments were analyzed when studying
the causes and effects of the UHI. Imhoff et al. (2010) analyzed various cities in the United
States and how the different biomes in each city affect the UHI effect in different ways. Using
13
land surface temperature (LST) data and impervious surface area (ISA) data, they conclude that
there is a positive correlation between the intensity of the Urban Heat island effect, the size of
the city, and the biome of the city (Imhoff et al. 2010). They also concluded that fraction ISA is a
good predictor of expected land surface temperatures for all continental U.S. cities in all biomes
except deserts and xeric shrublands (Imhoff et al. 2010). Tran et al. (2006) validates the findings
in Imhoff et al. (2010) by analyzing various major Asian cities in temperate and tropical climates
and found that the UHI intensities were positively correlated to population size of the cities. The
distinct climates found in each biome can influence the intensities of the UHI due to specific
climatic patterns.
2.2. Urban Heat Island Studies in Los Angeles
In 2012, as part of Assembly Bill 296, the California Environmental Protection Agency
(CalEPA) was tasked to develop an index to quantify and visualize the urban heat island effect
for various cities around California. Using this index, researchers and policy makers can quantify
the urban heat island effect into measurable values and can assess various goals set to mitigate
the effects of the urban heat island. This index is based on temperature, in the form of total
degree-hours. It is calculated by utilizing variables derived from remote sensing, such as surface
reflectance, topography, land use class, and temperature data. The UHII was calculated for each
census tract in total degree-hours units (DH) and DH per day averages, in Celsius (°C.hr/day)
(CalEPA 2015). The report described a large range of UHII values in different regions of
California and UHII values varied from 2 – 20°C.hr/day in smaller urban areas and 125°C.hr/day
or more in larger urban areas (CalEPA 2015). The report concluded that the cause of the large
14
variations in UHII values is due to the different microclimates and environments found in each
of the cities the report analyzed, such as topography.
One of the regions the CalEPA report analyzed was the Los Angeles Basin. It classified
the Los Angeles Basin area as an urban archipelago, which are continuous, large expanses of
urban areas, typically bounded by mountains or coastlines (CalEPA 2015). These urban
archipelagos act as one large continuous source of heat, rather than multiple pockets of heat that
are distributed throughout the city (CalEPA). The report found that the inland regions of the Los
Angeles Basin experience a higher UHII due to their distance from the coastline, which brings in
sea breezes to cool the region down. UHII values were found to be the highest close to the
mountain foothills, at the end of the downwind blowing inland (CalEPA). The models ran for the
report found an average air-temperature difference of 6-8°C between inland areas, such as San
Bernardino, and coastal cities found to the east of Los Angeles International Airport (LAX)
(CalEPA). The UHI effect in the Los Angeles basin is distributed more uniformly, due to the
urban archipelago effect causing the warming to be more uniform.
The various land uses within the Los Angeles metropolitan area also affect the intensities
of UHI. Most of the Los Angeles region is urbanized, with the distribution of UHI much more
equal compared to other cities (Roth et al. 1989). However, these UHIs can be detected with the
use remote sensing. In their study of Los Angeles, Roth et al. 1989 found that all the warm spots
identified as UHIs were found in heavily industrial areas and commercial centers, such as South
Norwalk and Commerce, which contains industrial buildings and railyards (Roth et al. 1989). In
another study, Dousset (1992) found similar areas exhibiting intense UHI effects, such as
Downtown Los Angeles, Vernon, and Whittier. This is due to the industrial surfaces’ ability to
absorb solar radiation and warm both the air and the adjacent residential areas. In The two
15
studies result both conclude that heavily urbanized and cities containing industrial development
showed the most intense UHI effect within the Los Angeles region.
2.3. Standardizing Urban Heat Island Studies
The UHI is one of the most studied climatic phenomena, and researchers have calculated
and quantified this effect using many different methods. However, there are concerns regarding
the accuracy and thoroughness of many UHI studies that have been presented over the years. A
study done by Iain Douglas Stewart in 2011 categorized these problems into three different
groups: (1) Studies which do not explicitly define UHI and the measurements to quantify UHI;
(2) Studies which did not define the difference between urban and rural areas; and (3)
Inconsistent methodology (Stewart 2011). The lack of clarity and consistency in many UHI
studies mean that it is often difficult to find studies without flaw, as the flaws often question the
overall credibility of the study. Stewart’s analysis in UHI studies found that around one-half of
the UHI studies sampled can be considered credible with complete and competent methodology
and reporting (Stewart 2011). Solving this issue would require a methodology which would
standardize the criteria for defining urban and rural areas and quantifying the UHI.
The use of a Local Climate Zone (LCZ) map can standardize the methodology for
defining urban and rural areas and the quantification of the UHI. One of the first standardized
methods for urban climatic studies is the LCZs. It involves classifying a region based on 10
different building types and 7 different land cover types. The UHI is then quantified as the UHI
magnitude, the calculation of temperature differences between each LCZ (Stewart & Oke 2012).
This approach is used as Stewart & Oke (2012) argue that the urban–rural temperature difference
poorly represents the UHI effect as it cannot differentiate between the different surface and
16
exposure characteristics between the different sample sites (Stewart & Oke 2012). It is often
difficult to determine whether the land cover or the buildings found in the urban and rural areas
is the main cause of the UHI in a region. Using the LCZ classification system is a relatively
simple and inexpensive method to standardize urban climate reporting in any region around the
world.
The LCZ model has been successfully utilized for Hong Kong, China, to visualize the
built environment. The studies done by Wang et al. 2018 and Zheng et al. 2018 involved three
major steps: (1) Perform sensitivity tests of spatial scale and geolocation of LCZ raster
framework, (2) Develop a set of urban morphology/land cover analysis maps and the LCZ
classification map at city scale, and (3) Analyze spatial distribution pattern of LCZ classes and
quantify urban morphology characteristics for typical LCZ sites (Wang et al. 2018; Zheng et al.
2018). The resulting LCZ classes were grouped based on a combination of land cover types, such
as bare soil and dense tree cover, and land use types, such as agricultural, shrub, or vacant
development land. The authors of the two studies produced the LCZ map for Hong Kong to
show the urban morphologies of the city and how each LCZ classification can affect the local
temperatures.
2.4. Urban Climatic Map
The urban climatic map integrates urban climatic factors and planning policies and
considerations together to visualize the urban morphology and its impacts on the local climate
(Baumüller et al. 1992). The urban climatic map brings another dimension to the traditional LCZ
maps by including the topographical, wind, and urban planning aspect to the maps. This includes
land use policies, and information on urban infrastructure such as building height, materials, and
17
footprints, which would affect the local microclimates. The urban climatic map is designed to
integrate urban climatology and urban development, two separate disciplines which would
benefit users, such as urban planners. These planners can use this map to help make informed
urban development decisions for a city.
Building the urban climatic map requires various datasets in order to create an overlay
showing the different urban climates of the region. These components include thermal load,
dynamic potential, and wind data. The thermal load aspect involves the region’s ability to absorb
and retain heat, while the dynamic aspect describes the region’s ability for air ventilation, to
move the heat away from the urbanized areas. While there are many different approaches to
building an urban climatic map, the datasets used typically involve an analysis in local air and
wind patterns. Wind patterns can include existing and potential air paths, location of barrier
effects by buildings, vegetation, or terrain, and various green space information and other urban
planning datasets (Ng & Ren 2015). The other datasets used for the urban climatic map are
selected based on their impacts to the thermal load and dynamic potential in each city.
The urban climatic map’s advantages include the ability to combine climatic knowledge
and urban planning recommendations into specific planning policies for a city, such as for master
plans, zoning plans, and land use plans (Ng and Ren 2015, 24). Problematic and sensitive areas
can be visualized in the urban climatic map, which allows relevant government agencies to take
action to mitigate such issues, such as the UHI effect. The simplicity of also makes the urban
climatic map easy to communicate the idea of urban climates and other relevant information to
the public.
However, there are limitations to the urban climatic map, especially when considering the
datasets that the urban climatic map is composed of. As the urban climatic map is
18
multidisciplinary in nature, it is important to incorporate the expertise and knowledge of urban
climatologists in order to fully evaluate the accuracy of the urban climatic map (Ng and Ren
2015, 24). The climate classes categorized by the urban climatic map are also based on land use,
and not necessarily an accurate representation of the microclimate at a place in time (Ng and Ren
2015, 24). Urban climate studies must also be done to supplement the urban climatic map in
order to accurately define each climate class in the study region. The urban climatic map are
focused mainly for urban planning purposes, and as such, must also be presented with planners
and land developers’ interests in mind.
Urban climate maps have been used as a visualization tool to aid urban climatologists and
urban planners around the world. It is not a recent concept and has been successfully utilized by
various cities since the late 20
th
century. German researcher Professor Kar Knoch was one of the
first researchers to utilize the urban climatic map. He proposed a series of urban climatic maps to
be utilized for city planning purposes (Ng and Ren 2015, 11). Later, various German researchers
in various cities began to conduct the Stadtklima, or ‘Urban Climate’ in German, as an attempt to
mitigate air pollution problems, especially in the Ruhr Valley region in western Germany.
Government documents and booklets, such as the one created for the city of Stuttgart, contains
information as to the variables that were used to create the UC maps for the city, including
climatic, topographic, and urban planning datasets (Baumüller et al. 1992). The urban climatic
map has since been used in cities of different sizes, such as Yokohama, Japan, Manchester,
United Kingdom, and Göteborg, Sweden.
Based on the work done by German urban climatologists, Ng et al. (2008) developed an
urban climatic map for Hong Kong. Research was done to select the datasets which were
appropriate for representing the urban climate of Hong Kong. Ng et al. (2008) utilized datasets
19
including building volume, topography/elevation, green space, ground coverage, natural
landscape, and proximity to openness to build the urban climatic map. These variables were
ranked based on whether the variables had an intensifying or mitigating effect on the UHI. Once
the variables were ranked, the raster datasets were overlaid to create a map with various integer
calculations, each representing a different urban climatic class. The different classes are then
classified based on their impact on thermal comfort for the region’s population. Urban planners
can then utilize the different classes visualized by the urban climatic map to create various UHI
mitigation plans in Hong Kong.
The flexibility of the urban climatic map allows other cities to include other datasets
which can best represent the thermal load and dynamic potentials of the specific cities. The study
by Ren et al. (2013) created an urban climatic map for Kaohsiung, Taiwan. Similar to the urban
climatic map for Hong Kong, the Kaohsiung urban climatic map is based on urban climate
studies done by German researchers, with the objective of classifying the climatic variations of
the city. However, Ren et al. differed from the urban climatic map of Hong Kong as it included a
separate layer for waterbodies for the urban climatic map. Waterbodies, such as lakes, ponds, and
the coastline found in the city has a cooling effect on its surroundings. The ability to include
other variables and datasets into the urban climatic map makes it a flexible tool for urban climate
analysis, as the map can be customized based on features of the city, such as the specific
topography, climate, or urban development.
2.5. Urban Heat Island Mitigation Strategies
In addition to studies on the causes and effects of UHI, there have been many studies on
UHI mitigation strategies in cities. Populations living in these cities affected by UHI are
considered vulnerable, as these individuals may not have the capability or access to resources to
20
adapt and mitigate the harmful effects of UHI. The abnormally warm temperatures can lead to
heat stroke and heat exhaustion, especially in the elderly and young, and can lead to increased
rates of mortality (Kovats and Hajat, 2008). Lemonsu et al. (2015) identified how UHI can be
influenced and intensified based on urban planning policies, and other urban expansion strategies
based on simulations through an urban canopy model in Paris, France. Using the simulations,
Lemonsu et al. (2015) found that city densification appears to increase the effects of UHI more
than urban sprawl growth. However, the study did note that there were many different factors
that could be analyzed individually from the simulations. Lemonsu et al. (2015) also concluded it
is also difficult to understand which factor may have the most influence in determining the
intensity of UHI. Understanding how planning policies, such as land use can unintentionally
affect the population by influencing the UHI effects is a key factor in mitigation strategies.
Government agencies, companies, and the public can utilize the knowledge provided in
studies to create strategies and policies to mitigate UHI. Recognizing certain parts of the
population as particularly vulnerable to UHI would allow organizations, such as the government,
to provide them with resources and aid. This can be especially true in poorer neighborhoods,
which experience a correlation between higher percentages of poor and minority inhabitants and
high heat stress exposure (Harlan et al. 2006). Reid et al. (2009) also lists other variables which
can identify vulnerable populations within a city. This include analyzing demographic,
socioeconomic, land cover, preexisting health condition, and air conditioning prevalence data
within a certain region. Using this information, researchers can visualize the locations of
population most susceptible to urban heat island effects. (Reid et al. 2009). Government agencies
can also utilize this information to provide policies or inform the public on practices to help
mitigate the effects of UHI.
21
The U.S. Government Environmental Protection Agency (EPA) has published reports to
the public with information on UHI and strategies that can be taken to mitigate UHI. This
include the use of green and cool roofs and pavements, encouraging the use of urban trees and
vegetation, public outreach activities, and at the policy level, where specific tree and landscaping
ordinances, building codes, zoning codes, and design guidelines to mitigate the effects of UHI.
(U.S. EPA 2012). Santamouris (2019) groups the mitigation technologies into four categories:
Reflective and Shading, Greenery, Heat Dissipation, and Anthropogenic Heat Reduction
technologies. Each of these categories aim to mitigate different aspects of UHI, such as the heat
absorption or heat dissipation aspects.
Studies have also analyzed the effectiveness of these technologies, such as the study done
by Solecki et al. (2005), which reviewed the technologies and approaches taken by two New
Jersey cities to mitigate the effects of UHI. Solecki et al. (2005) analyzed the effectiveness of
adding urban vegetation and reflective roofs as part of the mitigation strategies in New Jersey.
They found that the urban vegetation and reflective roofs are effective at reducing the effects of
the UHI. However, less affluent neighborhoods often do not have the resources or open space
available to utilize urban vegetation or reflective roofing (Solecki et al. 2005). Urban planners
and other stakeholders would need to analyze various factors to determine the most effective
strategies and technologies for UHI mitigation.
UHI mitigation efforts in Los Angeles would also need to account for air pollution, a
significant health hazard to the population of Los Angeles. Rosenfeld et al. (1998) analyzes a
program in the city to reduce the UHI and indirectly save on energy costs. Reroofing and
repaving to lighter colors in the city can reduce the UHI by as much as 3°C (Rosenfeld et al.
1998). As part of the 100 Resilient Cities program funded by the Rockefeller foundation, Los
22
Angeles produced a document which outlines plans and strategies to build resilience against
future physical, social, and economic challenges. The resiliency plan for Los Angeles proposes
basic planning guidelines which may help mitigate the effects of UHI. This can be used in
conjunction with current best use practices and recommendations for specific plan areas to build
a UHI mitigation plan that would best help each specific neighborhood.
23
Chapter 3 – Methods
The methods chapter describes the methods and data sources used to complete the urban
climatic map model for Los Angeles. The chapter lists and describes the datasets used to build
the urban climatic map, and its usage and significance in the model. The research design section
documents the process of building the model for Los Angeles based on urban climate maps
completed for other cities. It also includes a brief description of the different urban climatic
classes found in the urban climatic map, and the methodology used to create a regression model
to validate and compare the urban climatic map with another urban heat island model completed
for the Los Angeles region.
3.1. Data Sources
The extent of the urban climatic map of Los Angeles covers the area within city
boundaries, which encompasses an area of approximately 1240 square kilometers (479 square
miles), according to records from the City of Los Angeles, Bureau of Engineering. The map is
composed of classified pixels with data aggregated to 10 square meter resolution, with each pixel
covering a 10-meter by 10-meter area. As a result, the majority of the datasets used for the Los
Angeles urban climatic map are derived from the 2017 Los Angeles Region Imagery Acquisition
Consortium (LARIAC) orthoimagery, captured in the spring of 2017, to take advantage of the
high spatial resolution of the imagery. The 10 square meter resolution would result in a model
that can visualize the thermal loads and dynamic potentials at a finer scale.
24
3.1.1. Los Angeles Region Imagery Acquisition Consortium Datasets
The Los Angeles Region Imagery Acquisition Consortium (LARIAC) is composed of
many government agencies in the greater Los Angeles area with the goal of acquiring high-
resolution imagery for the county. EagleView, the contractor for the LARIAC project, captures
imagery approximately every three years, with each capture being a new iteration in the program.
There are currently five iterations of the LARIAC program, with LARIAC-5 orthoimagery
captured in 2017. As of Spring 2020, imagery collection for LARIAC-6 is currently being
captured. For the Los Angeles urban climatic map, the orthoimagery and datasets used is based
on the 2017 LARIAC datasets. The imagery dataset is publicly available online through the City
of Los Angeles ArcGIS Geohub and the Los Angeles County’s GIS portal.
The LARIAC orthoimagery covers the entire extent of Los Angeles County, with a
spatial resolution of four inches in urban areas and one-foot resolution in forested and vegetated
areas. The orthoimagery is composed of four spectral bands: red, green, blue, and near infrared,
and is visualized using a natural color scheme. The orthoimagery is delivered and visualized in
GIS software as tiles, with each tile covering an area of approximately ¼ square mile in urban
areas and 1 square mile in forested and vegetated areas. The orthoimagery is delivered in the
State Plane Coordinate System, NAD83, California, Zone V, U.S. Survey Feet (0405). As the
urban climatic map for Los Angeles shares the same projection, the orthoimagery and the
derived datasets do not have to be projected to a different coordinate system. The orthoimagery
has been checked for quality assurance and is reported to have achieved an accuracy of 1.35 feet
with a 95% confidence interval. Other datasets have also been processed and derived from the
LARIAC orthoimagery, including layers for building footprint and the elevation contours.
The building footprint layer is a polygon feature class layer which shows the perimeters
of the building structure on a property. Fields include the elevation of the building’s base and the
25
height of the building, derived from the LIDAR data collected from the same LARIAC program.
Structures with distinguishable roof features and specific heights are also included in the
building footprint layer attributes.
The digital elevation model (DEM) is the other dataset derived from the LARIAC
orthoimagery. The dataset is available as a contour line layer on the Los Angeles Geohub and is
provided at the one-foot, ten-feet, or twenty-feet contour interval. The ArcGIS geoprocessing
tool “Topo to Raster” is used to convert the contour data to a raster DEM. The resulting DEM
has a spatial resolution of around 9 square feet and has a dimension of 51,686 pixels across by
76,959 pixels in height.
3.1.2. Government Datasets
In addition to the datasets from the LARIAC-5 program, datasets from government
sources are also used for the Los Angeles urban climatic map. This includes political boundary
layers, the general plan land use, parks layer, the land cover data, and wind data. Excluding the
wind dataset and the land cover dataset, all the layers from government sources can be accessed
by the public through the Los Angeles ArcGIS Geohub and are available as polygon feature class
layers. The land cover dataset is a raster layer derived from the 2017 LARIAC orthoimagery and
was created using a supervised classification tool to classify the imagery into eight classification
categories: Bare Soil, Buildings, Grass and Shrubs, Roads and Railroads, Tall Shrubs, Tree
Canopy, Water, and Other Paved Surfaces. Table 1 is a summary of these datasets.
26
Table 1: Summary of Government Sourced Datasets
Dataset Source Data Type
Los Angeles County
Boundary
Los Angeles County
Department of Public Works
Polygon feature class
Los Angeles City Boundary Los Angeles Bureau of
Engineering
Polygon feature class
Los Angeles General Plan
Land Use
Los Angeles Department of
City Planning
Polygon feature class
Los Angeles Parks Layer Los Angeles Department of
Recreation and Parks & Los
Angeles Bureau of
Engineering
Polygon feature class
Los Angeles Land Cover
Layer
Los Angeles Department of
City Planning
TIF Raster File
Hourly wind speed and
Hourly wind direction
National Oceanic and
Atmospheric National
Centers for Environmental
Information
Comma Separated Values
File
Land-Based Wind Speed
80m
United States Department of
Energy, National Renewable
Energy Labs
TIF Raster File
The wind dataset is collected from the National Oceanic and Atmospheric Administration
(NOAA) National Centers for Environmental Information (NCEI). Under the NCEI program, the
public can access various climate records, including the Local Climatological Dataset (LCD).
The LCD contains hourly, daily, and monthly summaries of climate observations at various
weather stations across the United States. These observations include temperature, precipitation,
air pressure, and wind data. For the urban climatic map, hourly wind data will be gathered and
summarized by season to create wind roses visualizing the prevailing wind conditions during the
season. Two fields will be used to construct the wind rose: wind speed and wind direction. The
wind speed is the velocity of wind observed given in miles per hour, while wind direction
27
measures the direction in which the wind is blowing from at any given location. This is measured
in directional degrees, with 360° representing north directions and 180° representing south
directions. Data recorded with 0° in the directional fields indicate calm wind conditions, which
typically describe wind speeds of less than 3 miles per hour.
The Land-Based Wind Speed at 100 meters is provided by the United States Department
of Energy, National Renewable Energy Laboratory (NREL). The dataset visualizes projected
mean annual wind speed found in a specific area. The original raster layer has a spatial resolution
of 2.5 meters and represents wind velocities projected at a 100-meter height. According to the
NREL, interpolation methods were used to predict wind speed potential values at a finer
resolution. The dataset is visualized as a web-based map, called the “Wind Prospector,” which
estimates wind speed potentials across the country.
3.2. Research Design
The Los Angeles urban climatic map methodology emulates the methodology completed
for the urban climatic map of Hong Kong by Ng and Ren (2012). The Los Angeles urban
climatic map uses the same layers used in the Hong Kong urban climatic map to analyze the
thermal loads and dynamic potentials of the city. These layers include building volume,
topography, green spaces, ground coverage, natural landscapes, and proximity to openness. The
Los Angeles urban climatic map also uses the same classification values as the Hong Kong urban
climatic map to classify the impacts each layer has on the thermal loads and dynamic potentials.
Ng and Ren (2012) created an additional wind layer using meteorological data and wind models
to analyze the prevailing wind and air ventilation conditions for Hong Kong. Due to the time
frame of the project, meteorological data is used to construct wind arrows to visualize prevailing
28
wind conditions for Los Angeles. Figure 2 visualizes the methodology based on the work done
by Ng and Ren (2012).
Figure 2: Los Angeles urban climatic map methodology
29
3.2.1. Thermal Load Map:
The thermal load aspect of the urban climatic map describes the region’s ability to absorb
and retain heat within an urbanized area. According to Evans and de Schiller (1996), these
thermal loads are the variations in urban air temperatures due to the surfaces and urban forms
found in the region. Urban forms, which include anthropogenic structures such as commercial
buildings and paved asphalt surfaces, can influence the amount of solar radiation that is absorbed
and retained within a given area. Three separate layers are created as components to the thermal
load layer, which include building volume, topography, and green space.
3.2.1.1. Building Volume
The building volume layer is used as a measurement of the potential heat that a specific
building can retain. A larger building volume results in a higher potential heat capacity, which
can be represented within the thermal load layer. The heights of tall urban structures can also
obstruct the open sky from the ground level. This reduces the ability for the heat to dissipate into
the atmosphere and causes the surface to remain warmer for a longer period of time (Ng and
Ren, 2012). The building heights and density of buildings restrict air flow and ventilation, which
affect the ability for wind to transport warm air away and allow cooler air to flow into the area.
The building footprint layer is derived from the 2017 LARIAC dataset, which contains a
height field and area field in feet and square feet, respectively. The two fields represent the
height and the floor area of a given building footprint. Multiplying the values in the two fields
together will result in the building volume, given in cubic feet. The resulting building volumes is
then converted to cubic meters, to match the units used in Ng and Ren (2012). In the Hong Kong
urban climatic map, building volume percentages were calculated and used for classification in
30
the following steps. This is done by dividing a given building volume value with the highest
building volume value found in the entire city of Los Angeles and converted to a percentage
value.
The building footprint layer excludes areas with impervious surfaces. It is necessary to
include a land cover type layer to identify impervious areas which do not contain any buildings,
as impervious surfaces would have a much lower thermal load impact than areas containing an
urban structure. A query is used to select impervious areas and the attributes will be exported as
its own separate layer. Afterwards, the impervious area layer is merged with the building
footprint layer to create an output layer containing both attributes. As the selected impervious
surface areas do not contain buildings, a value of zero will be assigned to the building volume
and building volume percentage fields. The ‘CODE’ field found in the output layer will be
populated with either the “Building” or “Impervious Surface” attribute, to represent the type of
anthropogenic cover found at the location.
The feature polygon layer is then be converted to a raster and aggregated to a resolution
of 100 square meters, based on the building volume field. Using the reclassify tool, the raster
layer will then be classified based on the percentage building volume in each aggregated pixel.
Areas within city boundaries and containing no building volume data are assigned a
classification value of 0, while other building volume percentage values are assigned according
to the classification used in Ng and Ren (2012). Pixels with a higher building volume percentage
values are assigned a larger classification value, which represents a larger impact on the thermal
load. Figure 3 illustrates the methodology used to build the layer, while table 2 shows the six
different classifications used to classify the building volume percentage values.
31
Figure 3: Building Volume Layer Methodology
Table 2: Building Volume Thermal Load Classification (Taken from Ng and Ren 2012)
Thermal Load Building Volume Percentage
(percentage range)
Classification Value
Zero 0 (no building) 0
Very Low 0 (paved areas) 1
Low >0 – 4 2
Medium >4 - 10 3
High >10 - 25 4
Very High Greater than 25 5
32
3.2.1.2. Topography
The topography of a region is crucial to developing an urban climatic map. The elevation
of a region can affect the ambient temperatures of a given area, with higher elevations often
much cooler than areas at lower elevations. This is due to the phenomena known as the
environmental lapse rate, in which ambient air temperatures can change based on the altitude.
According to a study by Golany (1996) on the climatic impacts of urban design, the ambient air
temperature generally decreases by 1°C every 100m increase in elevation. Areas at lower
elevations experience warmer ambient temperatures, and as a result, are more susceptible to
more intense UHI effects. This is reflected in Ng and Ren (2012)’s classification, where lower
elevations are assigned a higher thermal load classification value.
Using the contour layer generated from the LARIAC LIDAR dataset, a digital elevation
model is created to represent the elevation for the city of Los Angeles. The digital elevation
model has a spatial resolution of 100 square meters, to match the resolution of the other layers of
the urban climatic map. As the original LARIAC dataset’s units are in feet, the elevation values
are converted to meters, to match the units used in the Hong Kong urban climatic map. The
minimum classification value is adjusted to account for Wilmington, Los Angeles, with a small
section of the city lying a few meters below sea level. Figure 4 illustrates the methodology for
the elevation layer, and table 3 shows the classification scheme used for the topography layer.
33
Figure 4: Elevation Layer Methodology
Table 3: Elevation Thermal Load Classification (Taken from Ng and Ren 2012)
Thermal Load Elevation (in meters) Classification Value
High >-10 - 50 0
Medium >50 - 100 -1
Low >100 – 200 -2
Very Low >200 -3
3.2.1.3. Vegetation
The third and final variable in the thermal load aspect is the vegetative cover layer. This
layer describes the amount of vegetative cover found in a location. Vegetation has a cooling
effect in urban areas and is often used by cities to reduce the urban heat generated by urban
structures. Through evapotranspiration, plants can absorb solar radiation and cool the
temperatures of heavily urbanized areas.
34
The vegetation of Los Angeles can be detected by calculating the NDVI using the
LARIAC orthoimagery. The LARIAC Orthoimagery contains four bands: Red, Green, Blue, and
Near-Infrared, two of which are used to calculate the NDVI. The equation for NDVI is given
below.
(1)
The results of the NDVI is on a scale between negative and positive one, where values
close to negative one indicate anthropogenic or bare ground cover, and values close to one
indicate dense vegetation. Once the NDVI values are calculated, the pixels will be aggregated to
a spatial resolution of 100 square meters and reclassified. NDVI values above 0.1 will be
classified as vegetative cover and values below 0.1 will be classified as non-vegetated cover.
Figure 5 illustrates the methodology to build the vegetation layer, and table 4 shows the NDVI
values and the impact of the vegetation on thermal load in urban areas.
Figure 5: Vegetation Layer Methodology
35
Table 4: Vegetation Thermal Load Classification (Taken from Ng and Ren 2012)
Thermal Load NDVI Values Classification Value
Negative (Cooling) Greater than or equal to 0.1 -1
Neutral Less than 0.1 0
3.2.2. Dynamic Potential Map:
The dynamic potential aspect analyzes the urban surfaces and features that affect the
wind through urbanized areas. The movement of air and the ability to ventilate air in an urban
setting are crucial when analyzing the ability for the urban area to allow wind to transport heat
away from the urbanized areas. The dynamic potential for Los Angeles will be composed of
three layers, which describe ground coverage, natural landscapes, and proximity to openness, as
defined by Ng and Ren (2012).
3.2.2.1. Ground Coverage
Similar to the effect on thermal load, tall buildings have a positive effect on urban
warming as they obstruct wind and air movement. This reduces the ability for the wind to
transport warmer air away from urban centers and prevents cooler air to flow in from open
spaces. The poor air ventilation creates a positive feedback loop, as the weaker air ventilation is
not strong enough to reverse the warming effects within the urbanized area (Tsang et al., 2012).
The study by Tsang et al. (2012) analyzed the effects of the shapes of the buildings, which can
reduce air ventilation in areas between such buildings using a wind model. While the building
density in Los Angeles is not as high as in other major cities, such as New York or Hong Kong,
there are pockets of dense urban development which can affect the area’s dynamic potential.
36
The ground coverage layer is derived from the 2017 LARIAC building footprint layer.
Using the merged layer created during the building footprint layer, a new field is created in the
building footprint layer and populated with a value of one, to indicate the presence of a building
or impervious surface at that location. The building footprints layer is converted to a raster based
on the cover field, with a spatial resolution of one square meter. The raster is then aggregated
into 100 square meter pixels, with the aggregated pixel value taking the average of the one
square meter pixels. These aggregated pixel values serve as the ground coverage percentage
values. Following Ng and Ren (2012)’s classification scheme, classification values are assigned
according to the dynamic potential and the ability for air to move through the urban area. Figure
6 illustrates the methodology for the ground coverage layer, and table 5 shows the three classes
for ground coverage.
Figure 6: Ground Coverage Layer Methodology
Table 5: Ground Coverage Dynamic Potential Classification (From Ng and Ren 2012)
Air Ventilation Potential Ground Coverage (in percent) Classification Value
High 0 – 30 -2
Medium >30 – 50 -1
Low >50 0
37
3.2.2.2. Natural Landscapes
Natural landscapes, as defined in Ng and Ren (2012), describe vegetative cover which
can alter the air flow and wind patterns with the city. Oke (1988) has analyzed and listed the
aerodynamic properties of various features found within an urban environment, including trees.
Depending on the size in a forest environment, trees have a surface roughness similar to
buildings, and have the ability to restrict air flow. The Ng and Ren study calibrated the green
space raster to assign a positive classification value to trees to describe the lower dynamic
potential that trees have.
Identifying the trees is a similar process to identifying the vegetative cover in the thermal
load section. However, the difference is that the threshold value is much higher to distinguish
trees from other types of vegetation, such as grass and shrubs. Values above the threshold will be
reassigned a value of one, while values below the threshold will be assigned a value of zero. This
describes the reduced dynamic potential that trees have, with its ability to restrict airflow in a
given area. Finally, the pixels will be aggregated to the 100 square meter resolution to make it
consistent with the other layers for the urban climatic map. Figure 7 shows the methodology for
the natural landscape layer, and table 6 shows the classification for the green space dynamic
potential as defined in Ng and Ren (2012).
38
Figure 7: Natural Landscape Layer Methodology
Table 6: Green Space Dynamic Potential Classification (Taken from Ng and Ren 2012)
Dynamic Potential NDVI Values Classification Value
Low Greater than or equal to 0.6 1
High Less than 0.6 0
3.2.2.3. Proximity to Openness
The final component of the dynamic potential aspect is the proximity to openness. Open
spaces are typically the source of cool air that blows through urbanized areas to ventilate the
area. This is due to lack of urban features which obstruct air flow and allow open spaces to
maintain cooler temperatures than developed urban areas. This layer is composed of three
different proximity calculations: distances measured from the coastline, an open space/park
within the city, and slope, which has the potential to increase the movement of air and the
cooling effect in the city.
39
The first open space layer calculates the distance away from the coastline, where the sea
is a source of cooler air that has the potential to ventilate urban areas. Buffering the coastline
data multiple times at various distances is done to visualize the diminishing cooling effect sea
breezes have as it moves inland. Ng and Ren (2012) has set the buffering values at 70 meters and
140 meters away from the coastline. Once the buffers are created, a field is added to the layer,
and a classification value is assigned to this field based on the distance the point is from the
coastline. In addition to the coastline proximity buffer, ground coverage is also considered when
determining the dynamic potential of the buffered regions. The final layer is a raster layer
aggregated to 100 square meters to match the spatial resolution of the urban climatic map. The
methodology for this layer is visualized as a flowchart in figure 8, while table 7 shows the
different dynamic potential values based on the different distances from the coastline.
Figure 8: Proximity to Waterbodies Layer Methodology
40
Table 7: Distance from Coastline Dynamic Potential Classification (From Ng and Ren 2012)
Dynamic Potential Distance from Coastline (in
feet)
Classification Value
High Close to coastline (0-70m)
and low ground coverage
(<30%)
-2
Medium Close to coastline (0-70m)
with medium ground
coverage (30-50%) or
medium distance from
coastline (70-140m) with low
ground coverage (<30%)
-1
Low Far from coastline (>140m)
or high ground coverage
(>50%) or medium distance
from coastline (70-140m)
with medium ground
coverage (30-50%)
0
The second open space layer calculates the distance away from the open spaces within the
city, such as parks. The parks data can be obtained from the parks dataset, while the open spaces
can be obtained from the general plan land use layer from the Los Angeles Department of City
Planning. A buffer of 100 meters around the parks and open spaces found within the city is done
to simulate the movement of cooler air from open spaces into the built environment. The open
space areas and the buffered areas are then merged and converted to a raster, with each pixel
covering an area of 100 square meters. Identified pixels containing parks and open spaces are
assigned a classification value of -1. The methodology is visualized in figure 9, and table 8
shows the open space dynamic potential classifications.
41
Figure 9: Proximity to Open Spaces Layer Methodology
Table 8: Open Space Dynamic Potential Classification (Taken from Ng and Ren 2012)
Raster Benefits from Open Space Classification Value
Yes -1
No 0
The final component in the open space layer is the slope from the digital elevation model.
In normal circumstances, steep vegetated terrain has the potential to increase wind velocity. The
Santa Ana winds in Southern California are a type of wind event that moves air from the Great
Basin towards the southern California region. The surrounding mountain range, the Transverse
Range, funnels the wind through the canyons, increasing the wind velocity and the dynamic
potential to ventilate urban areas. As a result, canyons with steeper slopes will have a higher
dynamic potential for air ventilation, which will produce a cooling effect on urban heat,
regardless of temperature.
42
Using the digital elevation model, the slope is calculated and aggregated to 100 square
meter pixels. Pixels with a slope of 40% or more are classified with a value of -1 and used as a
threshold value to determine whether the area the raster covers is benefiting from the dynamic
potential of the slopes. The vegetation layer is used in conjunction with the slope layer to
identify vegetated slopes. If there is vegetated cover overlaid on the pixels classified with a slope
greater than 40%, then the classification value of -1 will be kept. Otherwise, the pixel’s
classification value will be changed to zero. The methodology is illustrated in figure 10, and
table 9 shows the dynamic potential classification for slopes.
Figure 10: Proximity to Open Spaces Layer Methodology
Table 9: Slope Dynamic Potential Classification (Taken from Ng and Ren 2012)
Vegetated slopes of 40% and above Classification Value
Yes -1
No 0
43
3.2.3. Wind Data
Wind data is a highly complex variable and is difficult to accurately model in an urban
climatic map. However, throughout its existence, researchers have taken many different
approaches to model the necessary wind data and incorporate wind information into the urban
climatic map. In early urban climatic maps, German researchers utilized wind roses to identify
the prevailing wind conditions and air ventilation zones within the city (Baumüller et al. 1992).
As a result, later urban climatic map studies have used the wind rose to visualize the key wind
patterns within the study area, using the early methodologies as a standard. Ng and Ren (2012)
built upon this work to model an accurate wind environment for Hong Kong. Using a
combination of weather station datasets and the MM5 mesoscale meteorological model, Ng and
Ren (2012) predicted prevailing wind conditions and identified air ventilation zones within the
city.
For the Los Angeles urban climatic map, the wind dataset is taken from the National
Oceanic and Atmospheric (NOAA) National Centers for Environmental Information (NCEI).
NOAA has various weather stations located around Los Angeles, usually at airfields or large
open spaces where environmental factors such as people or tall buildings would not have a
significant impact on the temperature reading. The dataset contains hourly, daily, and monthly
averages for climate information, such as temperature, precipitation, humidity, air pressure, and
wind speed and direction.
Due to the different wind conditions that are present in the Los Angeles area, it is
necessary to distinguish between each of these wind conditions for the urban climatic map.
Prevailing winds, sea breezes, and seasonal wind conditions, such as the Santa Ana winds are
analyzed to understand wind patterns that affect the urban climate. The NCEI dataset is divided
into two datasets, one analyzing wind data for the months of June to August, and the second
44
analyzing wind data for the months of October to January. This is done to capture the wind data
under normal summer season conditions when the UHI is strongest, and the seasonal Santa Ana
wind conditions. Table 10 identifies the weather stations and their locations where the wind
speed and wind direction data are collected for the Los Angeles urban climatic map.
Table 10: NOAA NCEI Weather Stations
Name/Location Station ID Latitude Longitude
Burbank-Glendale-
Pasadena Airport
WBAN:23152 34.20056 -118.3575
Hawthorne Municipal
Airport
WBAN:03167 33.92278 -118.33417
Long Beach Airport WBAN:23129 33.8116 -118.1463
Los Angeles
Downtown, USC
WBAN:93134 34.0236 -118.2911
Los Angeles
International Airport
WBAN:23174 33.938 -118.3888
Los Angeles
Whiteman Airport
WBAN:53130 34.25917 -118.41333
San Gabriel Valley
Airport
WBAN:03165 34.08333 -118.03333
Santa Monica
Municipal Airport
WBAN:93197 34.01583 -118.45139
Van Nuys Airport
WBAN:23130
34.20972 -118.48917
Zamperini Field
WBAN:03174
33.80338 -118.33961
45
Wind roses are constructed to average the hourly wind information dataset gathered from
the weather stations under the NCEI program and visualize the prevailing wind conditions for the
month. A Python script is used to parse the wind data for each weather station and will output a
wind rose summarizing the prevailing wind conditions during the summer months and the Santa
Ana wind months. Using the wind roses, wind arrows representing the prevailing wind
conditions are added as two separate maps, one to show the wind conditions during summer
months and the other map showing wind conditions during the Santa Ana wind months. Figure
11 shows the Python script used to generate the wind roses, and figure 12 shows the wind rose
based on data from the summer months of 2017 recorded at USC.
Figure 11: Example Wind Rose Python script
46
Figure 12: Summer 2017 USC Wind Rose in miles per hour
3.2.4. Urban Climatic Map
The final urban climatic map is created by overlaying the classified raster layers together,
while also including the wind arrows created from the wind data. Using the raster calculator tool,
the thermal load map and the dynamic potential map are created by adding the six classified
raster layers together. The thermal load map is the sum of the building volume, topography, and
vegetation raster layers, while the dynamic potential map is the sum of the ground coverage,
green space, and proximity to openness layers. Table 11 is a summary table of the urban climatic
variables, what the variable represents, the variable data sources, the variable weight value
ranges showing the range of values used to classify and weight each layer, and the variable’s
impact to the urban heat island.
47
Table 11: Urban Climate Variable Summary
Variable
Name
Variable Description Variable Sources Variable
Weights Value
Range
UHI
Impact
Building
Volume
Percentage of the
building volume
found in the city
LARIAC Building
Footprint Layer, City
Wide Land Cover
Layer
0, 1, 2, 3, 4, 5 Positive
Topography Elevation of the area LARIAC 2017 Contour
Dataset
-3, -2, -1, 0 Negative
Vegetation Areas containing
vegetation
LARIAC 2017
Orthoimagery
-1, 0 Negative
Ground
Coverage
Percentage of
ground that is
covered by an
impervious surface
or building
Building Volume Layer
(Derived from LARIAC
Building Footprint)
-2, -1, 0 Negative
Natural
Landscape
Areas with clusters
of trees
LARIAC 2017
Orthoimagery
0, 1 Positive
Proximity
to Openness
Distance from open
areas, such as the
coast, parks, open
spaces, and
vegetated slopes
Los Angeles County
Boundary Layer, City
Wide Land Use Layer,
Ground Coverage Data
(Derived from LARIAC
Building Footprint
Layer), LARIAC
Contour Dataset
-4, -3, -2, -1, 0 Negative
48
The resulting raster classes are then reclassified and grouped together to create eight
urban climate classes as described in Ng and Ren (2012). These eight classes are based on the
different thermal loads and dynamic potentials that is expected for the area. These urban climate
classifications also describe the urban heat island potential and impacts of the area. Table 12
summarizes the eight different urban climate classification types based on the classifications for
the Hong Kong urban climate map by Ng and Ren (2012), and the corresponding colors used to
symbolize the classes.
49
Table 12: Urban Climate Classification Descriptions
Urban
Climate
Class
Class Symbol Classification Name Classification Description
1
Moderate Negative
Thermal Load and
Good Dynamic
Potential
High elevation areas and steep
vegetated slopes.
2
Some Negative
Thermal Load and
Good Dynamic
Potential
Areas covered by natural vegetation,
greenery, including hilly slopes and
coastal areas.
3
Low Thermal Load
and Good Dynamic
Potential
Areas containing spaced out
development with little ground
coverage. Includes open areas within
the city and sparse development in
coastal areas.
4
Some Thermal
Load and Some
Dynamic Potential
Areas containing low to medium
building volumes in an open,
developed environment. These include
open spaces found in between
buildings.
5
Moderate Thermal
Load and Some
Dynamic Potential
Areas containing medium building
volumes located in low elevation
inland areas.
6
Moderately High
Thermal Load and
Low Dynamic
Potential
Areas consisting of medium to high
building volumes found in low
elevation areas and found usually in
developed areas with little green
space.
7
High Thermal Load
and Low Dynamic
Potential
Areas containing a majority of high
building volumes located in well-
developed, low elevation areas.
8
Very High Thermal
Load and Low
Dynamic Potential
Areas containing very high and
compact building volumes with very
limited open areas. Very little air
ventilation due to the obstruction
caused by tall buildings.
50
3.2.5. Validation and Linear Regression
Validating the results of the urban climatic map involves comparing the classification
results of the urban climatic map with another map which classifies the urban heat island effect.
The CalEPA’s Urban Heat Island Index is used as a reference to verify the accuracy of the urban
climatic map. Both the CalEPA Urban Heat Island Index and the urban climatic map describe the
potential warming and cooling in a specific region. This similarity allows the Urban Heat Island
Index to be grouped into the same number of categories as the urban climatic map in order to
make a comparison between the two models.
A linear regression model is used to explain the relationship between the CalEPA Urban
Heat Island Index and the urban climatic map for Los Angeles. However, the CalEPA model
utilizes a mesoscale meteorological model, which includes wind and temperature as a key input
in determining the urban heat island index, a variable that is not present in the urban climatic
map. As a result, an Ordinary Least Squares (OLS) regression model is used to determine any
missing variables which is missing from the urban climatic map when comparing it with the
CalEPA Urban Heat Island Index model. While the CalEPA model uses temperatures based on
the Degree Day units to represent the warming and cooling of a region, temperature
measurements have variable effects which can differ based on region. As a result, the only
significant variable that can be tested in the OLS regression model is the wind speed class
variable.
51
Chapter 4 - Results
The results chapter presents the components created for the urban climatic map as well as
how they are used to build Los Angeles’ urban climatic map. The chapter is organized by each of
the components of the urban climatic map, the thermal load variables and the dynamic potential
variables. The results are visualized through maps describing the classifications for each of the
variables in the urban climatic map. Finally, the results of the linear regression model to identify
the fitness of the urban climatic model with the CalEPA’s urban heat island index model.
4.1. Thermal Load Map
The thermal load map is a measurement of the potential heat the region can absorb. It is
based on building volume, topography, and vegetation of the area. building volume variable is
created from a combination of the building footprint layer and impervious attributes selected
from the citywide land cover layer. Figures 13, 14, and 15 show the final building volume,
topography, and vegetation layers that are used as the components to produce the thermal load
map. Figure 16 is the final thermal load map which would be one of the final components for the
urban climatic map.
52
Figure 13: Building Volume
53
Figure 14: Topography
54
Figure 15: Vegetation
55
Figure 16: Thermal Load Map
56
4.2. Dynamic Potential Map
The dynamic potential map is a representation of how well air can move throughout a
region. It is based on the ground coverage, natural landscapes which impede air movement such
as trees, and proximity to openness, as described in the Hong Kong urban climatic map. The
ground coverage and natural landscape variables are shown in figures 17 and 18, respectively.
The proximity to openness itself is composed of three other variables, which include proximity
to waterfront, proximity to open spaces, and vegetated slopes. Figures 19, 20, and 21 show the
components of the proximity of openness variable, shown in figure 22. Figure 23 shows the final
dynamic potential map which is used as the component to build the draft urban climatic map.
57
Figure 17: Ground Coverage
58
Figure 18: Natural Landscape Cover
59
Figure 19: Proximity to Waterfront
60
Figure 20: Proximity to Open Spaces
61
Figure 21: Vegetated Slopes
62
Figure 22: Proximity to Openness
63
Figure 23: Dynamic Potential Map
64
4.3. Urban Climatic Map without the Wind Layer
The draft urban climatic map is produced from combining the thermal load map and the
dynamic potential map. This process creates a raster with values ranging from -10 to 5. Figure 24
shows the reclassified urban climatic map into the eight urban climate classes.
65
Figure 24: Draft Urban Climatic Map
66
4.4. Validation/Regression Map
The linear regression model to compare the fitness of the urban climatic. The first
Ordinary Least Squares model is used to identify the fitness of the urban climatic map classes
with the degree day values by census tract from the CalEPA urban heat island index. The
resulting R
2
value for this model was 0.033050. The low R
2
value indicates a large discrepancy
between the CalEPA urban heat island index model and the urban climatic map model. Figure 25
shows the results of the Ordinary Least Squares model, with calculations representing model
significance and statistical biases. The regression equation to model the best fit between the two
variables is also given below.
Figure 25: Ordinary Least Squares result for CalEPA UHII and Aggregated Urban Climatic
Classes
Regression Equation:
(1) CalEPA UHII = -2.756451(Aggregated Urban Climatic Map) + 19.403551
67
The second Ordinary Least Squares model uses the projected wind speed values from the
National Renewable Energy Laboratory to test whether the wind variable is a significant key
variable that is a missing component in the linear regression model. The resulting R
2
value for
the second model including the wind variable was 0.314852. While the resulting R
2
value
indicates that the aggregated urban climatic map with the wind variable is still not a best fit, the
result indicates that the wind variable is possibly one of the key exploratory variables that is
missing between the aggregated urban climatic map when comparing the model with the CalEPA
UHII model. Figure 26 shows the results of the second Ordinary Least Squares model, which
includes the projected wind speed classes. The regression equation to model the best fit line is
also given below.
Figure 26: Ordinary Least Squares result for CalEPA UHII, Aggregated Urban Climatic Classes,
and Projected Wind Speed Values
Regression Equation:
(2) CalEPA UHII = -7,539121(Aggregated Urban Climatic Class) - 16,025709(Wind Speed Classification) + 33,739969
68
The residuals from the second Ordinary Least Squares model are also mapped as an
output from the Ordinary Least Squares tool. The residuals visualize the difference between the
actual data points and the projected data points that is plotted as a result of the linear regression
equation. These results can be used to verify the accuracy of a model or dataset and any potential
missing variables with a reference dataset. Figure 27 shows the residuals generated from the
aggregated urban climatic map and projected wind dataset with the reference CalEPA UHII
model.
69
Figure 27: Ordinary Least Squares Standardized Residuals
70
4.5. Urban Climatic Map with Wind Layer and Prevailing Wind Information
The final urban climatic map takes into account the prevailing wind information from the
National Centers for Environmental Information and the projected wind speed information from
the National Renewable Energy Laboratory. The prevailing wind information is presented as
wind arrows above the weather station that the wind observations were recorded. Figures 28 and
29 show the wind roses generated by the Python script for each of the weather stations in the Los
Angeles city vicinity, separated by the summer months and the Santa Ana wind months. Figure
30 shows the locations of the NCEI weather stations where the wind speeds and directions were
observed. Figures 31 and 32 show the wind roses created from the wind dataset observed at each
of the weather stations shown in figure 30. The summer months are defined as the months of
June to August and the Santa Ana wind months are defined from October to December. The
projected wind speed information is reclassified into classes based on the wind velocity
categories presented in the wind prospector web application. Figure 33 shows the final urban
climatic map for Los Angeles including the summer month wind information, while figure 34
shows the final urban climatic map for Los Angeles with the Santa Ana month wind conditions.
71
Figure 28: Summer Months Wind Roses
72
Figure 29: Santa Ana Wind Months Wind Roses
73
Figure 30: Weather Station Locations
74
Figure 31: Summer Month Wind Roses
75
Figure 32: Santa Ana Months Wind Roses
76
Figure 33: Final Urban Climatic Map without Wind Layer
77
Figure 34: Final Urban Climatic Map with Wind Layer
78
Chapter 5 - Discussion and Conclusion
The purpose of this project is to develop an urban climatic map for the city of Los
Angeles and test its accuracy and effectiveness as an urban heat island visualization model by
comparing it with another urban heat island model for the region. Referencing the classification
scheme developed by Ng and Ren (2012) for the urban climatic map of Hong Kong, the urban
climatic map was developed by classifying variables related to the thermal load and dynamic
potential aspects, such as building volume, vegetative cover, ground cover, and distance from
open areas. This chapter highlights observations made about the results of the urban climatic
map, how the urban climatic map compares to the results of other urban heat island studies done
in the Los Angeles area, addresses limitations, and proposes possible future work to improve the
urban climatic map for Los Angeles and in other similar regions around Southern California.
5.1. Observations
The majority of the lower urban climate classes are distributed in the hillside areas of Los
Angeles, such as the Santa Monica Mountains, the San Gabriel Mountains, and the areas of
Griffith Park. These results are expected as the low urban climatic classes describe open spaces
and other areas with little to no urban development. Similarly, we see higher urban climatic
classes in more heavily urbanized areas, such as Downtown Los Angeles. The taller buildings
and low vegetative cover, combined with mostly impervious surfaces would increase the thermal
load and reduce the dynamic potential of the area, which can be represented through the higher
urban climatic classes.
The urban climatic classes also reveal that on average, various residential communities,
such as San Pedro and South Los Angeles, are classified much higher than other more heavily
urbanized places containing commercial and industrial development. Typically, these
79
neighborhoods experience higher temperatures and possible greater urban heat island effect due
to the topography, where nearby ranges block the sea breeze from blowing into these
communities. The Palos Verdes Ranges prevent westward sea breezes from blowing into San
Pedro, while the Baldwin Hills restrict sea breezes from blowing into the South Los Angeles
communities to the east. The final urban climatic map including the wind speed classification
layer visualizes these effects with higher classification values.
The elevation appears to have the biggest weight when looking at the complete urban
climatic map, with urban climatic class changes defined by the classification scheme used in the
topography layer. be weighted heavily. When comparing residential areas of different elevations,
higher urban climatic classes are found in low elevations, such as in South Los Angeles, and
lower classes are found in the San Fernando Valley, at a higher elevation. Areas of Wilmington
are also classified with higher urban climatic classes, possibly due to its sea level elevation.
The creation of wind roses was done to replicate the work done by Ng and Ren (2012) on
their urban climatic map, where they used wind speed and direction observations to illustrate
prevailing wind information, and wind models to show air ventilation paths for the region. Due
to the unique wind conditions of Los Angeles, it is more appropriate to show the wind roses,
rather than just wind arrows, in order to fully understand the prevailing wind conditions of the
region. As a result, to make the wind roses more noticeable on a map, the methodology was
changed to show the wind speed and direction information as its own individual map, rather than
integrate prevailing wind arrows within the urban climatic map itself.
Analyzing the results of the urban climatic map with other urban heat island studies of
Los Angeles reveals several differences. The main difference that can be observed between the
Los Angeles urban climatic map and other Los Angeles urban heat island studies is the locations
80
of where the most intense urban heat island events are occurring. In older studies, more intense
urban heat islands were found to be focused around industrial and commercial areas, as shown in
the works of Roth et al. (1989) and Dousset (1992). The materials used in industrial buildings
and impervious surfaces kept temperatures warmer than residential areas within the same region.
The results of the urban climatic map agree with a portion of these findings, where heavy
commercial and industrial areas of Downtown Los Angeles and the communities of Wilmington
and Harbor Gateway have a much higher urban climatic class than surrounding neighborhoods.
However, it is not possible to compare the results of other neighborhoods identified in the Roth
et al. and Dousset studies, as they are separate cities and neighborhoods outside the Los Angeles
city boundaries and are not included in the Los Angeles urban climatic map.
More recent studies, such as the one completed by Hulley et al. (2019) analyzes
neighborhood and city vulnerabilities to the urban heat island effect and analyzed the land
surface temperatures during the day and night. Using the National Aeronautics and Space
Administration’s (NASA) new Ecosystem Spaceborne Thermal Radiometer Experiment on
Space Station (ECOSTRESS), the researchers reveal that areas containing a large amount of
materials with high heat capacity, such as concrete, asphalt and waterbodies were able to absorb
large amounts of heat during the day and reradiate it during the night, which increases the overall
temperatures and the urban heat islands within these areas (Hulley et al., 2019). These areas
include the various airports scattered around the city, the port area, and valley areas, such as
areas in the San Fernando and San Gabriel Valley.
While the results of the urban climatic map and Hulley et al. (2019) both find that the
port areas and the industrial areas of Downtown Los Angeles experience much more intense
urban heat islands, proved by the higher temperatures detected by ECOSTRESS in Hulley et al.
81
(2019) and the higher urban climatic classes for these areas from the Los Angeles urban climatic
map. However, the urban climatic map fails to classify most of the San Fernando Valley with
higher urban climatic classes, when compared to the high temperatures detected in the San
Fernando Valley by the ECOSTRESS program. The urban climatic map also classifies much of
the residential neighborhoods of South Los Angeles with very high urban climatic classes, where
there are not any high the ECOSTRESS program does not. These two issues can be explained by
the elevation possibly having a much higher weight when creating the urban climatic map for
Los Angeles and is discussed further in the limitations section.
5.2. Limitations
While the groundwork for the Hong Kong urban climatic map presented by Ng and Ren
in their 2012 study can be easily replicated, the methodology can only be used to guide the basis
of developing urban climatic maps for other cities, such as Los Angeles. The complexities of the
urban climatic map, the different input datasets, and different variable classifications and
weightings would lead to inaccurate urban climatic modeling when the same model is applied to
different cities around the world. For example, the urban climates used to describe the compact,
dense building forms in a humid city such as Hong Kong would not accurately translate to
describing the urban climates of a semi-arid, urban sprawl city such as Los Angeles. Variables
such as wind are also difficult to model and include in the urban climatic map, due to its
complexity and how wind interacts and move through the environment. It is important to
understand the issues and inaccuracies that may arise when creating the urban climatic map for
Los Angeles.
The major limitation of the Los Angeles urban climatic map is the weighting and
classification used to classify each layer. As this project uses the classification scheme from the
82
Hong Kong urban climatic map done by Ng and Ren (2012), it cannot accurately represent the
urban climate of the Los Angeles region. Hong Kong is characterized by tall, highly dense
building forms, while Los Angeles is characterized by having a more uniform, medium to low
density urban environment. For example, the classification used is tailored to accurately model
the high rise, dense building forms of Hong Kong, but it may not accurately represent the
uniform low to medium building densities of Los Angeles. These differences would mean that
due to the values used for classification, some layers are weighted more heavily than others.
The weighting limitation can be observed by looking at the current system used to
classify and weight each of the layers in the urban climatic map. Some variables, such as the
building volume variables, are classified with values ranging from 0 to 5, while the vegetation
variable for the thermal load aspect is only classified using a value of -1 and 0. The classification
range used in the building volume layer is much higher than the other layers, which potentially
leads to a greater impact on the final urban climatic class value when all the layers are added
together using raster calculator. The values found in the building volume layer can shift a
specific pixel’s final urban climatic class number, where on the other hand, the vegetation layer
classification value of -1 would generally not change or alter any of the pixel’s urban climatic
class number. This limitation is an issue that must be addressed in future urban climatic map
methodologies, where layers have different weights on the overall urban climatic map based on
how many individual values were assigned as part of the classification process.
The classification of the building volume layer used for the Hong Kong urban climatic
map is not suitable to describe the building volumes found in Los Angeles. Ng and Ren (2012)
utilized sky view factors to assign classification values for the building volume layer. Sky View
Factor describes how much the visible sky can be viewed at a measured point, provided as a
83
ratio. This method is one of the more common methods of studying urban morphology and is
often used to describe a three-dimensional built environment feature as a two-dimensional
variable to be used in urban models (Middel et al. 2018). The sky view factor can also be
correlated with urban heat island intensity, as shown in Oke (1982). This analysis would be
necessary to correctly assign classification values to the building volume percentages, so that the
classifications can accurately describe the built environments of Los Angeles and be weighted
correctly in relation to the other layers in the urban climatic map.
Calculating the vegetation layers for both the thermal load and dynamic potential layers
utilized NDVI to determine the presence of vegetation in the area. However, the NDVI is a
relatively simple index which is a representation of the “greenness” of an object. This means that
any object that can reflect the same red, green, and near-infrared wavelengths as the amount
reflected from vegetation would result in the same NDVI values calculated. A study by Huete
and Jackson (1988) found that soil influences can be as significant as atmospheric effects on
remotely sensed indices and would need to be corrected in order to calculate accurate vegetation
values. Improving the vegetation layer for the Los Angeles urban climatic map can involve using
other vegetation indices, such as the Soil-Adjusted Vegetation Index (SAVI) and the Modified
Soil-Adjusted Vegetation Index (MSAVI) developed by Qi et al. (1994), in order to reduce the
influence of soils in vegetative cover calculations.
Topography appears to have the most influence on the Los Angeles urban climatic map.
The classification used by Ng and Ren (2012) are calibrated for Hong Kong, as most of the built
environment is at sea level, along the coasts. Hong Kong’s urban development is concentrated
around the coastal regions and the lower elevations, where the terrain is flat to build upon. The
classification used by Ng and Ren (2012) reflects this. Los Angeles, on the other hand, has urban
84
development several hundred meters above sea level. This would result in areas, such as the San
Fernando Valley, having negative classification values assigned using the Hong Kong urban
climatic map classification scheme, and underestimate the thermal load of the region. This can be
observed when comparing the results of the Los Angeles urban climatic map and how the low
urban climatic classes do not match the results from the Hulley et al. (2019) study. The results of
the ECOSTRESS land surface temperature experiment shows higher temperatures in the San
Fernando Valley, compared to the rest of the city. However, from the Los Angeles urban climatic
map, due to the classification of the elevation, the urban climatic classes suggest that the lower
elevations of South Los Angeles cause the urban climatic classes to be higher, while areas where
we expect much higher susceptibility to the urban heat island are assigned lower classification
values due to the valley’s elevation of a few hundred meters above sea level.
The elevation range in Hong Kong is also much smaller compared to the elevation range
found in the city of Los Angeles. The lowest point in Hong Kong is at sea level, along the
coastlines, while the highest point is Tai Mo Shan, rising 957 meters above sea level. On the
other hand, the lowest point in Los Angeles is the Wilmington neighborhood, found several
meters below sea level, while the peak of Mount Lukens is the highest point in the city, rising
more than 1540 meters above sea level. More research must be done to develop a more
appropriate classification scheme and have possible weights assigned to the layer to accurately
describe the elevation ranges of Los Angeles and its impact on the thermal load.
Wind conditions are not accurately represented in the Los Angeles urban climatic map. In
the Hong Kong urban climatic map, Ng and Ren (2012) performed various wind model tests, to
model possible air ventilation paths to understand how wind could move through the region.
These variables are considered to explain where the urban heat island effect can intensify and
85
where this effect may be negated as a result of the wind. The wind data used for the Los Angeles
urban climatic map is a projected wind speed map, which shows the potential wind speeds, and
therefore, not an accurate model of the wind speeds that have been observed during the time
period.
The geography of Hong Kong allows sea breezes to affect the majority of the urban
development found within the region. These effects can be represented by both the wind
conditions and the proximity to coastlines layer. Los Angeles, on the other hand, only contains
very few coastal communities which benefits from the sea breeze, such as the Westside
communities of Venice and Santa Monica. San Pedro, on the other hand, does not receive as
much of a benefit from the sea breeze, as the Palos Verdes Hills block onshore flow to cool the
region. For the Los Angeles urban climatic map, the proximity to coastlines would need to be
adjusted or combined with the wind data to develop more accurate weights for classification.
The wind roses generated from the Python script show the wind speeds and the
frequencies in which they occur. This was done to identify the prevailing wind speed and
direction of the region. However, the wind roses do not distinguish between wind patterns during
the day and night. This is an issue, as the Los Angeles area typically experiences onshore flow
during the day, and offshore flow during the night. In the wind roses, the reversing wind patterns
can sometimes appear to have the same frequencies in multiple directions, meaning it can be
difficult to identify the prevailing wind directions just from one wind rose at each location. This
issue can be solved by separating daytime wind observations and nighttime observations and
creating separate wind roses for each condition.
The effects of Santa Ana wind conditions are not modeled by the Los Angeles urban
climatic map using the methodology for the Hong Kong urban climatic map. As a result, the Los
86
Angeles urban climatic map may not represent the effects on the urban heat island during these
months. Temperatures often rise during the Santa Ana wind months while the wind strength can
have the potential to increase the dynamic potential of the region. More research will need to be
done to identify the effects and see which effects are the strongest during the Santa Ana months.
Due to the coverage and extent of the datasets used, the Los Angeles urban climatic map
covers only areas within the Los Angeles city boundaries. The unique shape of the city results in
the exclusion of many neighborhoods along the coast and inland communities where there is
more industrial urban development. Cities that border Los Angeles may contain elements that
can affect the overall thermal load and dynamic potential, and possibly urban heat island effects
along the boundaries of Los Angeles. Many other urban heat island studies have analyzed Los
Angeles as a region, including analyzing neighborhoods surrounding the city boundaries. As
shown in the observations, it was not possible to validate the results of the urban climatic map
with the studies done by Roth et al. (1989) and Dousset (1992), as they identified neighborhoods
outside the city of Los Angeles. This issue can be resolved in future studies by analyzing a larger
area, such as including the surrounding neighborhoods, or even the entire Los Angeles county to
calculate the urban climatic classes for the city.
The linear regression model completed for this project compares how well the Los
Angeles urban climatic map and potential missing variables fit with the CalEPA’s Urban Heat
Island Index, the established study used as the reference for validation. The Los Angeles urban
climatic model along with the NREL’s projected wind speed data resulted in a R
2
value of
approximately 0.31, which indicates that the two variables can account for 31% of the variability
between the CalEPA model and the urban climatic map. However, the low R
2
value indicates
there are many missing variables that have not been accounted for. These variables can include
87
meteorological or population variables used in the CalEPA UHII model. While the urban
climatic map is not a temperature-based classification, temperature can also be a key missing
variable that can account for the missing 69% variability between the two models.
5.3. Future Work
Future work on the Los Angeles urban climatic map involves addressing the limitations
of the map and the inclusion of other variables used in the CalEPA UHII model, along with other
urban climatic maps developed in the past. Further work would focus on differentiating between
urban and rural areas, accounting for missing variables, and accurately assigning weights and
classifications of variables used in the urban climatic map for Los Angeles. This can be
accomplished through analyzing Köppen climate classifications between the cities and utilizing
meteorological and climate models to analyze short-term weather conditions and long-term
climate. Comparison by individual variables in the regression model would also find which
individual variables have a much higher influence and greater weight in the city for the urban
climatic map. The wind layer used in the Los Angeles urban climatic map makes it difficult to
accurately model the wind conditions and its effect on the urban climates within the region.
Future work would also propose methods and other workflows to accurately model the complex
wind conditions in Los Angeles. Finally, the urban climatic map must be comprehensible to
stakeholders such as urban planners, who will ultimately develop policies, best use practices, or
recommendations to mitigate the effects of the urban heat island.
It is important to differentiate urban and rural areas within the city when discussion the
urban climate and the urban heat island. The urban heat island is often defined as the temperature
increase and warming observed in an urbanized setting. As a result, the urban climatic map for
Los Angeles treats the region as one homogenous built environment. Future work on the urban
88
climatic map would segment these areas further, where areas of Los Angeles can be classified as
urban or rural. One approach to differentiating between urban and rural areas is using Local
Climate Zone classification. According to Stewart and Oke (2006), these local climate zones are
classified based on regions of uniform surface cover, structure, material, and human activity. By
weighting the urban climate classes based on the local climate zones proposed by Stewart and
Oke, the urban climatic map can differentiate between urban and rural areas and apply the urban
climatic classes accordingly.
The weighting of each layer as a result of the classification process must be addressed,
which results in some layers of the urban climatic map that have a disproportionately higher
weight than other layers. Under normal circumstances, the seven variables used in the final urban
climatic map with the wind layer, each variable would be weighted equally at approximately
14.3%. In future work on the Los Angeles urban climatic map, the weighting can be adjusted for
each layer based on classification range used in each layer. As an example, the building volume
layer can be reweighted to match the classification values used in other layers in the urban
climatic map. Instead of assigning integer classification values from 0 to 5, the reweighted values
would then be assigned decimal values between 0 to 1, to match the classification values of -1
and 0 used in the vegetation layer. Since the final urban climatic map is a reclassified raster
based on the results of a raster calculator process, it would not be an issue to use decimals or
non-integer values in the classification. The urban climatic classes used in the map are the result
of a reclassification of a range of values resulting from the raster calculator process into eight
different classes.
The weighting and classification of the urban climatic map layers can also be customized
by climate classification. When looking at potential future studies where much larger regions
89
cover multiple Köppen climate classifications, it is important to understand the potential effects
that each climate classification can have on the urban heat island effect. Under the Köppen
climate classification scheme, Hong Kong’s climate is described as humid subtropical, while Los
Angeles has a Mediterranean climate, typically characterized by dry conditions. The two
climates are significant enough that the weights and classifications must be adjusted to account
for these differences, in order to accurately represent the regional climate’s impact to the urban
heat island. These climatic differences are not limited to the regional scale, but to the local scale
as well. Microclimates also exist within the Los Angeles area, with some of the coastal
communities experiencing much cooler temperatures than areas up in the San Fernando valley,
where the warmest temperatures in the entire city are often observed. In addition to customizing
weights by climate classification, future work on the urban climatic map for Los Angeles would
also focus on the microclimatic variations observed in the city.
Missing variables must be accounted for to fully develop an accurate model describing
the urban climate of the region. One example is the additional variables used in the CalEPA
UHII model that are not present in the Los Angeles urban climatic map, including population
density and other meteorological and climatological variables such as temperature, solar
radiation, and surface albedo. Urban climatic maps such as the one developed for Berlin,
Germany, and Arnhem, Netherlands utilize air temperatures during different times of the day as
an input variable for their analysis. Regional variables specific to a region should also be
considered to model certain characteristics. For Los Angeles, this could include analyzing air
pollution and its impact to the urban heat island effect and other urban developments.
In addition to identifying and including missing variables between the urban climatic map
and the CalEPA UHII model, work must also be done to compare each of the individual
90
variables used in both models. Using the ordinary least squares model, we can identify the
missing exploratory variables in more detail, and identify which variables are more significant to
Los Angeles, compared to the variables used in the Hong Kong urban climatic map. A one to one
comparison of the variables would be able to reveal variables which have more significance to
the Los Angeles region and identify variables which do not have a significant impact in the
region. This information would ultimately help with the reweighting and reclassification of the
layers for future urban climatic maps of Los Angeles.
One of the biggest limitations of the Los Angeles urban climatic map is the wind
variable. The complexity of wind makes it difficult to accurately model wind patterns, especially
for Los Angeles, which experiences daily onshore and offshore flow, and the seasonal Santa Ana
winds. Future work could improve on the wind data used for the current Los Angeles urban
climatic map by using higher spatial and temporal resolution data. Wind data interpolation from
weather station observations is another option that can be taken. This can also help distinguish
between day and nighttime wind conditions and between summer months and Santa Ana month
wind conditions.
The wind variable in the urban climatic map would also need to describe friction
surfaces, such as commercial buildings or terrain. And how moving air flows through the city,
interacting with these built environments. These friction surfaces can alter air flow, by changing
the wind speed or direction, which can impact how heat can be transported throughout the city
and impact the urban heat island. These changing wind observations can be visualized through
three-dimensional modeling of wind profiles and built environments which can help analyze the
horizontal and vertical wind vectors as they pass through buildings and other terrain.
Improvements in technology in the future can potentially allow for accurate, real-time wind
91
modeling, which can model the changes friction surfaces can have on the wind. This can lead to
quicker analysis and response on changing wind conditions within the microclimates.
Future work on the urban climatic map can utilize the improved technologies in the
future, which leads to improved data richness which can be drawn upon to improve future urban
climatic maps. This can include a higher frequency in which the data is collected, increased
resolution of the data, or better, more efficient ways of ground truthing and validating the data.
With the datasets available at the time of creating the initial urban climatic map for Los Angeles,
the classes are describing based on the capture date of the 2017 LARIAC orthoimagery and is a
snapshot of a point in time. Increased frequency of the observations and imagery capture can
allow for the urban climatic classes to be modelled in real time and allow for immediate
responses to issues, such as the increase in urban heat island within a specific microclimate in the
city in an hourly or daily basis.
Urban climatic maps are intended to be used by urban planners to aid in their decision
and policymaking to mitigate the effects of the urban heat island in their city. The urban climatic
map draw from variables and datasets that are familiar to urban planners and other policy
makers, such as land use, land cover, and building footprint information. It is important for urban
planners to understand the urban climate of the region and its classifications in the map that they
can easily understand, so the recommend policies and best use practices are applicable and
suitable for the region.
Urban planners have different goals and priorities when planning for their cities, and
while the urban climatic map proposes a somewhat standardized method for urban heat island
studies, the implementation of policy and how the urban climatic classes are analyzed will differ
by region. In many urban climatic map studies, such as the urban climatic map for Hong Kong,
92
an additional recommendations map is created based on the results of the urban climatic map.
The urban climatic planning recommendations map for Hong Kong recategorizes the urban
climate classes into five planning zones, with each zone contains a recommendation of action
relating to urban heat island mitigation strategies. Similar methodologies can be adapted for
other cities, such as Los Angeles, where planners can use the urban climate map to develop
policies and prioritize specific areas of Los Angeles to mitigate the most intense urban heat
island effects in the city.
5.4. Conclusion
The Los Angeles urban climatic map proposes a new methodology for urban heat island
studies, as a standardized way to model urban heat island in a region, by replicating the
methodology of other developed urban climatic maps. While there are many limitations as
discussed for the Los Angeles urban climatic map, the groundwork established by urban climate
researchers have been successful in other regions around the world, such as in Europe and Asia.
The quantification of urban classes can aid urban planners understand their cities’ urban
morphology, which can lead to better urban policies and recommendations, such as those which
can mitigate the effects of the urban heat island. While the methodology for the urban climatic
map will need to be customized based on the region being analyzed, these adjustments to the
variable classification and inclusion of other variables such as wind could improve the urban
climatic map for future areas in the region, with the purpose of replicating a similar methodology
for other cities in California. There are many environmental challenges that must be solved, and
the hope is that the methodology proposed in this project can form the groundwork for solutions
to mitigate such issues, such as climate change.
93
References
Arnfield, A. J. 2003. “Two decades of urban climate research: a review of turbulence, exchanges
of energy and water, and the urban heat island.” International Journal of Climatology,
23(1); 1–26.
Baumuller, J., U. Hoffmann, U. Reuter. 1992. Climate booklet for urban development, Ministry
of Economy Baden-Wuerttemberg (Wirtschaftsministerium), Environmental Protection
Department (Amt fur Umweltschutz).
California Environmental Protection Agency. 2015. Creating and Mapping an Urban Heat Island
Index for California. CalEPA / Altostratus Inc. Agreement No. 13-001.
https://calepa.ca.gov/wp-content/uploads/sites/6/2016/10/UrbanHeat-Report-Report.pdf
Dousset, Bénédict Marie. 1992. “Remote Sensing of Urban Microclimates and their Relationship
to Land Use: A Case Study of the Los Angeles Basin.” PhD diss., University of
California, Los Angeles.
Evans, J. M., & S. D. Schiller. 1996. “Application of microclimate studies in town planning: A
new capital city, an existing urban district and urban river front development.”
Atmospheric Environment, 30(3), 361–364.
Glickman, Todd S. 2000. Glossary of Meteorology (2nd ed.). Boston: American Meteorological
Society.
Golany, G.S. 1996. “Urban design morphology and thermal performance.” Atmospheric
Environment, 30(3), 455-465.
Harlan, S. L., A. J. Brazel, L. Prashad, W. L. Stefanov, & L. Larsen. 2006. “Neighborhood
microclimates and vulnerability to heat stress.” Social Science & Medicine, 63(11), 2847–
2863.
Huete, A. R., & Jackson, R. D. 1988. “Soil and atmosphere influences on the spectra of partial
canopies.” Remote Sensing of Environment, 25(1), 89–105.
Holmer, B., U. Postgård, & M. Eriksson. 2001. “Sky view factors in forest canopies calculated
with IDRISI.” Theoretical and Applied Climatology, 68(1-2), 33–40.
Imhoff, M. L., P. Zhang, R. E. Wolfe, & L. Bounoua. 2010. “Remote sensing of the urban heat
island effect across biomes in the continental USA.” Remote Sensing of Environment,
114(3), 504–513.
Kim, H. H. 1992. “Urban heat island.” International Journal of Remote Sensing, 13(12), 2319–
2336.
94
Kovats R. S. & H. Shakoor. 2008. “Heat Stress and Public Health: A Critical Review.” Annual
Review of Public Health. 29(1), 41-55.
Landsberg, H. E. 1981. The Urban Climate. New York: Academic Press.
Lemonsu, A., V. Viguié, M. Daniel, & V. Masson. 2015. “Vulnerability to heat waves: Impact of
urban expansion scenarios on urban heat island and heat stress in Paris (France).” Urban
Climate, 14(4), 586-605.
Manley, G. 1958. “On the frequency of snowfall in metropolitan England.” Quarterly Journal of
the Royal Meteorological Society, 84(359), 70–72.
Middel, A., J. Lukasczyk, R. Maciejewski, M. Demuzere, & M. Roth. 2018. “Sky View Factor
footprints for urban climate modeling.” Urban Climate, 25(1), 120-134.
Ng, Edward, and R. Chao, eds. 2015. The Urban Climatic Map: A Methodology for Sustainable
Urban Planning. Milton Park, Abringdon, Oxon OX14 4RN: Routledge.
Oke, T. R. 1982. “The energetic basis of the urban heat island.” Quarterly Journal of the Royal
Meteorological Society, 108(455), 1–24.
Oke, T. R. 1988. “The urban energy balance.” Progress in Physical Geography, 12(4), 471–508.
Oke, T. R. 1995. “The Heat Island of the Urban Boundary Layer: Characteristics, Causes and
Effects.” Wind Climate in Cities, 81–107.
Office of Los Angeles Mayor. March 2018. Resilient Los Angeles.
https://www.lamayor.org/sites/g/files/wph446/f/page/file/Resilient%20Los%20Angeles.p
df.
Parker, D. E. 2009. “Urban heat island effects on estimates of observed climate change.” Wiley
Interdisciplinary Reviews: Climate Change, 1(1), 123–133.
Peel, M. C., B. L. Finlayson, B. L., & T. A. McMahon. 2007. “Updated world map of the
Köppen-Geiger climate classification.” Hydrology and Earth System Sciences, 11, 1633–
1644.
Qi, J., A. Chehbouni, A. R. Huete, Y. H. Kerr, and S. Sorooshian. 1994. “A Modified Soil
Adjusted Vegetation Index.” Remote Sensing of Environment, 48, 119-126.
Reid, C. E., M. S. O’Neill, C. J. Gronlund, S. J. Brines, D. G. Brown, A. V. Diez-Roux, & J.
Schwartz. 2009. “Mapping Community Determinants of Heat Vulnerability.”
Environmental Health Perspectives, 117(11), 1730–1736.
Ren, C., K. L. Lau, K. P. Yiu, & E. Ng. 2013. “The application of urban climatic mapping to the
urban planning of high-density cities: The case of Kaohsiung, Taiwan.” Cities, 31, 1–16.
95
Rosenfeld, A. H., H. Akbari, J. J. Romm, & M. Pomerantz. 1998. “Cool communities: strategies
for heat island mitigation and smog reduction.” Energy and Buildings, 28(1), 51–62.
Roth, M., T. R. Oke, & W. J. Emery. 1989. “Satellite-derived urban heat islands from three
coastal cities and the utilization of such data in urban climatology.” International Journal
of Remote Sensing, 10(11), 1699–1720.
Santamouris, M. 2019. “Mitigating the Local Climatic Change and Fighting Urban
Vulnerability.” Minimizing Energy Consumption, Energy Poverty and Global and Local
Climate Change in the Built Environment: Innovating to Zero, 223–307.
Sarrat, C., A. Lemonsu, V. Masson, & D. Guedalia. 2006. “Impact of urban heat island on
regional atmospheric pollution.” Atmospheric Environment, 40(10), 1743–1758.
Solecki, W. D., C. Rosenzweig, L. Parshall, G. Pope, M. Clark, J. Cox, & M. Wiencke.
2005. “Mitigation of the heat island effect in urban New Jersey.” Environmental Hazards,
6(1), 39–49.
Stewart, I. D., & T. R. Oke. 2012. “Local Climate Zones for Urban Temperature Studies.”
Bulletin of the American Meteorological Society, 93(12), 1879–1900.
Stewart, I. D., & T. R. Oke, 2006. Methodological concerns surrounding the classification of
urban and rural climate stations to define urban heat island magnitude. Proc. Sixth Int.
Conf. on Urban Climate. Goteborg, Sweden, International Association for Urban
Climate, 431–444.
Taha, H. 1997. “Urban climates and heat islands: albedo, evapotranspiration, and anthropogenic
heat.” Energy and Buildings, 25(2), 99–103.
Tran, H., D. Uchihama, S. Ochi & Y. Yasuoka. 2006. “Assessment with satellite data of the
urban heat island effects in Asian mega cities.” International Journal of Applied Earth
Observation and Geoinformation, 8(1); 34–48.
Tsang, C. W., K. C. S. Kwok, & P. A. Hitchcock. 2012. “Wind tunnel study of pedestrian level
wind environment around tall buildings: Effects of building dimensions, separation and
podium.” Building and Environment, 49, 167–181.
U.S. Environmental Protection Agency. 2008. Reducing urban heat islands: Compendium of
strategies. Draft. https://www.epa.gov/heat-islands/heat-island-compendium.
Van Heerwaarden, C. C., & J. V. Guerau de Arellano, J. V. 2008. “Relative Humidity as an
Indicator for Cloud Formation over Heterogeneous Land Surfaces.” Journal of the
Atmospheric Sciences, 65(10), 3263–3277.
96
Wang, R., C. Ren, Y. Xu, K. K.- L. Lau, K, & Y. Shi. 2018. “Mapping the local climate zones of
urban areas by GIS-based and WUDAPT methods: A case study of Hong Kong.” Urban
Climate, 24, 567–576.
Weng, Q., D. Lu, & J. Schubring. 2004. “Estimation of land surface temperature–vegetation
abundance relationship for urban heat island studies.” Remote Sensing of Environment,
89(4), 467–483.
Westerling, A. L., D. R. Cayan, T. J. Brown, B. L. Hall, & L. G. Riddle. 2011. “Climate, Santa
Ana Winds and autumn wildfires in southern California.” EOS, 85(31), 289 – 300.
Zheng, Y., C. Ren, Y. Xu, R. Wang, J. Ho, K. Lau, & E. Ng. 2018. “GIS-based mapping of
Local Climate Zone in the high-density city of Hong Kong.” Urban Climate, 24, 419-
488.
97
5.5. Bibliography
City of Los Angeles, ENERGY AND ENVIRONMENT COMMITTEE. 2016. Energy and
Environment Committee Report relative to the formation of a Committee on Cooling and
Urban Heat Impacts, and related matters.
https://cityclerk.lacity.org/lacityclerkconnect/index.cfm?fa=ccfi.viewrecord&cfnumber=1
5-0198
Hagen, Bjoern & K. Pijawka. 2017. Sustainability for the 21st Century: Pathways,Programs,
and Policies. Dubuque, IA: Kendall Hunt Publishing.
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Wetland mapping and restoration decision making using remote sensing and spatial analysis: a case study at the Kawainui Marsh
PDF
Using GIS to perform a risk assessment for air-transmitted bioterrorism within San Diego County
PDF
Radar horizon estimation from monoscopic shadow photogrammetry of radar structures: a case study in the South China Sea
PDF
A comparative study of ground and satellite evapotranspiration models for southern California
PDF
Single & multi-family housing: the complex behind the simple: a kit of parts for minimally invasive interventions
PDF
Identifying and mitigating the effects of urban heat islands in California
PDF
Classifying ash cloud attributes of Eyjafjallajökull volcano, Iceland, using satellite remote sensing
PDF
A comparison of urban land cover change: a study of Pasadena and Inglewood, California, 1992‐2011
PDF
Mitigating the urban heat island effect: thermal performance of shade-tree planting in downtown Los Angeles
PDF
A spatial investigation of New York City's historical shoreline
PDF
Harnessing GIST-enabled resources in the classroom: developing a Story Map for use with secondary students
PDF
Urban areas and avian diversity: using citizen collected data to explore green spaces
PDF
Development and evaluation of a USV based mapping system for remote sensing of eelgrass extent in southern California
PDF
Beyond greenhouse gases and towards urban-scale climate mitigation: understanding the roles of black carbon aerosols and the urban heat island effect as local to regional radiative forcing agents
PDF
A comparison of pixel-based, object-based, deep learning, and data fusion image classification approaches in delineating urban tree canopy in City Terrace
PDF
The wildland-urban interface in Lassen County, California: a change analysis 2000–2015
PDF
Exploring the pernicious effects of redlining and discriminatory policies on an American city: a spatio-temporal case study of New York City
PDF
The impact of definition criteria on mapped wildland-urban interface: a case study for ten counties along the Oregon-California border
PDF
An exploratory spatial analysis of fire service and EMS accessibility in northeastern Illinois communities
PDF
Assessing the value of crowdsourced data in aiding first responders: a case study of the 2013 Boston Marathon
Asset Metadata
Creator
Lam, Bryan Gene
(author)
Core Title
Developing a replicable approach for the creation of urban climatic maps for urban heat island analysis: a case study for the city of Los Angeles, California
School
College of Letters, Arts and Sciences
Degree
Master of Science
Degree Program
Geographic Information Science and Technology
Publication Date
07/12/2020
Defense Date
03/25/2020
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
OAI-PMH Harvest,urban climate,urban heat island,Urban Planning
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Marx, Andrew (
committee chair
), Fleming, Steven (
committee member
), Ruddell, Darren (
committee member
)
Creator Email
bglam@usc.edu,bryla8127@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-327235
Unique identifier
UC11663768
Identifier
etd-LamBryanGe-8659.pdf (filename),usctheses-c89-327235 (legacy record id)
Legacy Identifier
etd-LamBryanGe-8659.pdf
Dmrecord
327235
Document Type
Thesis
Rights
Lam, Bryan Gene
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
urban climate
urban heat island