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
/
Historical temperature trends in Los Angeles County, California
(USC Thesis Other)
Historical temperature trends in Los Angeles County, California
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
HISTORICAL TEMPERATURE TRENDS IN LOS ANGELES COUNTY, CALIFORNIA
By
Dustin Dwayne Reed
A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
GEOGRAPHIC INFORMATION SCIENCE AND TECHNOLOGY
May 2015
Copyright 2015 Dustin Dwayne Reed
ii
DEDICATION
I would like to dedicate my thesis to my family and more specifically to my wife. Christa
has taken on a lot of extra burden because a lot of my time is spent on my schoolwork
and she has always supported me in completing my master’s degree. For this motivation
and support, I could not be more thankful. Also, a special thanks goes to my wonderful
and loving parents, Robert and Judy Reed. My Dad has always been an excellent role
model because he instilled within me a hard-work ethic, to always reach for your dreams,
and to always do the best that I can. Even though my Mom is not here to experience this
life achievement and her death leaves an unexplainable emptiness in my heart, it is
comforting to know that she is my angel watching over me and is always guiding me in
the right direction.
iii
ACKNOWLEDGMENTS
I extend my gratitude and appreciation to the University of Southern California for giving
me the opportunity to pursue my love of science and to continue my education in this
field of study. I owe my everlasting gratitude and appreciation to my committee chair
member and advisor Dr. Su Jin Lee, his intense questions and critiques taught me to think
“outside the box” and to keep digging deeper for improvement. Also, I would like to
thank Dr. Darren Ruddell and Dr. Robert Vos for their agreement to be a member of my
committee and for their constructive feedback.
A special thanks is owed to my best friend and wife, Christa, who has always made keep
pushing even when I was ready to throw in the towel, and to my parents who always put
my best interests before their needs. I love you three more than you will ever know!
iv
TABLE OF CONTENTS
Dedication ii
Acknowledgments iii
List of Tables vi
List of Figures viii
List of Abbreviations xii
Abstract xiv
Chapter One: Introduction 1
1.1 Measuring Climate Change 1
1.1.1 Climate Change across the World 3
1.1.2 Climate Change in the United States 6
1.1.3 Climate Change in Southern California 8
1.2 Research Questions 11
Chapter Two: Literature Review 13
2.1 Climate Change and its Impact on Temperature Changes 13
2.2 Threshold Temperature Analysis 16
2.3 Urban Heat Island Effect 19
Chapter Three: Data Sources and Methods 22
3.1 Description of Study Area 22
3.2 Data 26
3.2.1 Historical Surface Temperature 26
3.2.1.1 Daily Surface Temperature 26
3.2.1.2 Monthly Surface Temperature 27
3.3 Selection of Weather Stations 29
3.3.1 Weather Stations for Daily Surface Temperature 29
3.3.1.1 Characteristics of Daily Temperature Data at
Six Stations 31
3.3.2 Weather Stations for Monthly and Yearly Surface
Temperature 34
3.3.2.1 Weather Stations for Monthly Surface
Temperature 39
3.3.2.2 Weather Stations for Yearly Surface
Temperature 42
3.4 Analysis of the Climate Trend 44
3.4.1 Daily Temperature 44
3.4.2 Monthly Temperature 45
3.4.3 Yearly Temperature 45
v
Chapter Four: Results 46
4.1 Daily Temperature and Its Trend 46
4.1.1 Frost and Misery Day Annual and Decadal Statistics 46
4.1.1.1 Frost and Misery Day Linear Regression
Significance 55
4.1.1.2 Spatial Discontinuity of Three Weather Stations
and Their Trends 57
4.1.2 Heat Wave Decadal Thresholds 77
4.2 Monthly Temperature and Its Trend 84
4.2.1 Monthly Temperature Linear Regression Significance 92
4.3 Yearly Temperature and Its Trend 94
4.3.1 Recorded Temperature at the Twenty One Weather Stations 94
4.3.1.1 Recorded Temperature from 1931 to 1950 94
4.3.1.2 Recorded Temperature from 1951 to 2010 98
Chapter Five: Discussion & Conclusion 101
5.1 Extreme Temperature Threshold Observations 101
5.1.1 Weather Stations Experiencing an Accelerated
Warming Trend 101
5.1.2 Weather Station Experiencing an Accelerated
Cooling Trend 102
5.1.3 Weather Stations Experiencing a Modest
Warming Trend 104
5.2 Monthly Surface Temperature Observations 105
5.3 Yearly Surface Temperature Observations 107
5.4 Overall Surface Temperature Observations 108
5.5 Spatial Characteristics and Distribution of the Weather Stations 109
5.6 Current and Future Approaches 111
5.6.1 Current Study Advantages and Disadvantages 111
5.6.2 Future Study 112
References 113
vi
LIST OF TABLES
Table 1: Operating time period for the six weather stations containing daily
temperature data between 1931 and 2010 29
Table 2: Characteristics of the six weather stations containing daily temperature
data 33
Table 3: Operating time period for the 21 weather stations containing monthly
surface temperature data between 1931 and 2010 37
Table 4: Operating time period for the eight weather stations containing monthly
surface temperature data between 1931 and 2010 40
Table 5: Operating time period for the six weather stations containing yearly
surface temperature data between 1951 and 2010 42
Table 6: Operating time period for the 20 weather stations containing yearly
surface temperature data between 1931 and 1950 43
Table 7: Total number and annual mean of frost days and misery days measured
by decade at the Fairmont weather station for a total of 80 years (1931 to 2010) 47
Table 8: Total number and annual mean of frost days and misery days measured by
decade at the Los Angeles weather station for a total of 80 years (1931 to 2010) 48
Table 9: Total number and annual mean of frost days and misery days measured
by decade at the Palmdale weather station for a total of 80 years (1931 to 2010) 50
Table 10: Total number and annual mean of frost days and misery days measured
by decade at the Pasadena weather station for a total of 80 years (1931 to 2010) 51
Table 11: Total number and annual mean of frost days and misery days measured
by decade at the Sandberg weather station for a total of 80 years (1931 to 2010) 53
Table 12: Total number and annual mean of frost days and misery days measured
by decade at the UCLA weather station for a total of 80 years (1931 to 2010) 54
Table 13: Statistical significance characteristics for frost days at all six weather
stations 57
Table 14: Statistical significance characteristics for misery days at all six weather
stations 57
Table 15: Statistical significance characteristics for frost days at Los Angeles 64
vii
Table 16: Statistical significance characteristics for misery days at Los Angeles 65
Table 17: Statistical significance characteristics for frost days at Palmdale 71
Table 18: Statistical significance characteristics for misery days at Palmdale 72
Table 19: Statistical significance characteristics for frost days at Sandberg 76
Table 20: Statistical significance characteristics for misery days at Sandberg 77
Table 21: Heat wave characteristics by decade from daily temperature data for
the Fairmont weather station from 1931 to 2010 78
Table 22: Heat wave characteristics by decade from daily temperature data for
the Los Angeles weather station from 1931 to 2010 79
Table 23: Heat wave characteristics by decade from daily temperature data for
the Palmdale weather station from 1931 to 2010 80
Table 24: Heat wave characteristics by decade from daily temperature data for
the Pasadena weather station from 1931 to 2010 81
Table 25: Heat wave characteristics by decade from daily temperature data for
the Sandberg weather station from 1931 to 2010 82
Table 26: Heat wave characteristics by decade from daily temperature data for
the UCLA weather station from 1931 to 2010 83
Table 27: Statistical significance characteristics for monthly temperature at all
eight weather stations 93
viii
LIST OF FIGURES
Figure 1: Global surface temperature anomaly (°C) for four worldwide scientific
institutions from 1880-2020 3
Figure 2: Global surface temperature anomaly (°F) from 1901-2013 5
Figure 3: Global average surface temperature (°C) projections for 2020-2029
and 2090-2099 6
Figure 4: U.S. temperature anomaly (°F) from 1901-2013 across the contiguous
48 states 7
Figure 5: Annual average temperature trend for the State of California
(1895-2012) 9
Figure 6: Departure from average for mean temperature, minimum temperature,
and maximum temperature for the State of California (1895-2012) 10
Figure 7: Departure from average for mean temperature, minimum temperature,
and maximum temperature for the South Coast Region in the State of California 11
Figure 8: Los Angeles County and its location within the State of California 24
Figure 9: Distribution of 106 weather stations containing historical monthly
surface temperature across Los Angeles County, California 28
Figure 10: Spatial location of six weather stations with 80 years or more of
historical daily temperatures 30
Figure 11: Large-scale aerial imagery representation of six weather stations
across Los Angeles County 34
Figure 12: Selected 21 weather stations across Los Angeles County, California
for monthly surface temperature 38
Figure 13: Spatial location of eight weather stations with approximately 80 years
of historical monthly temperatures 41
Figure 14: Annual number of frost days and misery days using daily temperature
data at the Fairmont weather station from 1948 to 2010 47
Figure 15: Annual number of frost days and misery days using daily temperature
data at the Los Angeles weather station from 1931 to 2010 49
ix
Figure 16: Annual number of frost days and misery days using daily temperature
data at the Palmdale weather station from 1931 to 2010 50
Figure 17: Annual number of frost days and misery days using daily temperature
data at the Pasadena weather station from 1931 to 2010 52
Figure 18: Annual number of frost days and misery days using daily temperature
data at the Sandberg weather station from 1948 to 2010 53
Figure 19: Annual number of frost days and misery days using daily temperature
data at the UCLA weather station from 1931 to 2010 55
Figure 20: Annual number of frost days and misery days using daily temperature
data at the Los Angeles weather station from 1931 to 1939 59
Figure 21: Annual number of frost days and misery days using daily temperature
data at the Los Angeles weather station from 1940 to 1964 60
Figure 22: Annual number of frost days and misery days using daily temperature
data at the Los Angeles weather station from 1964 to 1985 61
Figure 23: Annual number of frost days and misery days using daily temperature
data at the Los Angeles weather station from 1985 to 1999 62
Figure 24: Annual number of frost days and misery days using daily temperature
data at the Los Angeles weather station from 1999 to 2007 63
Figure 25: Annual number of frost days and misery days using daily temperature
data at the Los Angeles weather station from 2007 to 2010 64
Figure 26: Annual number of frost days and misery days using daily temperature
data at the Palmdale weather station from 1931 to 1948 66
Figure 27: Annual number of frost days and misery days using daily temperature
data at the Palmdale weather station from 1948 to 1952 67
Figure 28: Annual number of frost days and misery days using daily temperature
data at the Palmdale weather station from 1952 to 1962 68
Figure 29: Annual number of frost days and misery days using daily temperature
data at the Palmdale weather station from 1962 to 1982 69
Figure 30: Annual number of frost days and misery days using daily temperature
data at the Palmdale weather station from 1982 to 1993 70
x
Figure 31: Annual number of frost days and misery days using daily temperature
data at the Palmdale weather station from 1993 to 2010 71
Figure 32: Annual number of frost days and misery days using daily temperature
data at the Sandberg weather station from 1948 to 1981 73
Figure 33: Annual number of frost days and misery days using daily temperature
data at the Sandberg weather station from 1982 to 1996 74
Figure 34: Annual number of frost days and misery days using daily temperature
data at the Sandberg weather station from 1996 to 2000 75
Figure 35: Annual number of frost days and misery days using daily temperature
data at the Sandberg weather station from 2000 to 2010 76
Figure 36: Monthly surface temperature trend at the Fairmont weather station
from 1931 to 2010 85
Figure 37: Monthly surface temperature trend at the Los Angeles International
Airport weather station from 1944 to 2010 86
Figure 38: Monthly surface temperature trend at the Palmdale weather station
from 1931 to 2010 87
Figure 39: Annual number of frost days and misery days using daily temperature
data at the Pasadena weather station from 1931 to 2010 88
Figure 40: Monthly surface temperature trend at the Sandberg weather station
from 1948 to 2010 89
Figure 41: Monthly surface temperature trend at the UCLA weather station
from 1933 to 2010 90
Figure 42: Monthly surface temperature trend at the USC weather station
from 1950 to 2010 91
Figure 43: Monthly surface temperature trend at the Woodland Hills weather
station from 1949 to 2010 92
Figure 44: Yearly surface temperature for the twenty weather stations in Los
Angeles County, California 96
Figure 45: Yearly summer surface temperature for the twenty weather stations in
Los Angeles County, California 97
xi
Figure 46: Yearly surface temperature for the six weather stations in Los Angeles
County, California 99
Figure 47: Yearly summer surface temperature for the six weather stations in Los
Angeles County, California 100
xii
LIST OF ABBREVIATIONS
C2ES Center for Climate and Energy Solutions
CNV California-Nevada Region
COOP National Weather Service Cooperative Network
CSV Comma-Separated Value
DWR Department of Water Resources
GISS NASA Goddard Institute for Space Studies
IPCC Intergovernmental Panel on Climate Change
JMA Japanese Meteorological Agency
km Kilometer
LAX Los Angeles International Airport
MOHC Met Office Hadley Centre/Climatic Research Unit
NASA National Aeronautics and Space Administration
NCDC National Climatic Data Center
NOAA National Oceanic and Atmospheric Administration
OEHHA Office of Environmental Health Hazard Assessment
PCM Parallel Climate Model
PRISM Parameter-elevation Regressions on Independent Slopes Model
R
2
Coefficient of Determination
RSS Remote Sensing Systems
T1 Temperature Threshold 1
T2 Temperature Threshold 2
xiii
UAH University of Alabama-Huntsville
UCLA University of California-Los Angeles
USC University of Southern California
USHCN US Historical Climate Network
WMO World Meteorological Organization
WRCC Western Regional Climate Center
xiv
ABSTRACT
Climate change is a global occurrence and is studied at multiple scales within Los
Angeles County, California. Determining the type of surface temperature trend across
Los Angeles County is best observed using historical daily, monthly, and yearly
temperature data. Each type of historical temperature data is analyzed for various
temperature and extreme temperature threshold trends: (1) thresholds of frost days
(minimum temperature ≤ 32°F), misery days (maximum temperature ≥ 90°F), and heat
wave events are examined at six weather stations; (2) type of linear trend is measured for
monthly surface temperature at eight weather stations; and (3) type of linear trend is
analyzed for yearly surface temperature and yearly summer surface temperature (July to
September) for twenty weather stations from 1931 to 1950 and six weather stations from
1951 to 2010.
This study’s major findings are (1) daily maximum and minimum surface
temperature show strong departures from normal conditions for threshold temperature
trends as Palmdale experiences an accelerated warming trend and Sandberg experiences
an accelerated cooling trend; (2) a variance in decadal heat wave thresholds exists at each
weather station for 80 years; (3) monthly mean surface temperature is a good source to
reflect seasonal temperature variations; and (4) yearly surface temperature is not
sufficient temperature data to track temperature trends. Analyzing surface temperature
trends is a tool for monitoring how climate change is impacting temperatures globally.
The following chapters include: (1) introduction is the motivation and research
questions; (2) literature review is previous studies on climate change and its impact on
temperature; (3) data and methods are data sources and the implementation of these
xv
sources; (4) results offer a detailed explanation and examples of the findings; (5)
discussion is an overview of the important findings; and (6) references are sources that
are cited within the manuscript.
1
CHAPTER ONE: INTRODUCTION
Numerous studies and agencies (International Panel on Climate Change 2007; National
Aeronautic and Space Administration (NASA) 2014a; United States Environmental
Protection Agency 2014; Office of Environmental Health Hazard Assessment 2014) have
investigated climate systems from the local to global scales of analysis. Climate data can
play a very important role in monitoring and predicting climate change by providing
valuable temperature measurements across the globe. These valuable temperature
measurements (observed, anomaly, and projected) serve as the primary source for
discovering how temperature is changing at various global locations. Chapter one
introduces examples of climate change at these various scales as well as population
growth and its influence. Also, this chapter states the research questions for the study at
hand.
1.1 Measuring Climate Change
Climate change and its effect on environmental, economic, and social issues is a
debated topic throughout the public and government domain. There are numerous
scientific organizations dedicated to understanding climate change and its effects. For
instance, the Pew Center on Global Climate Change (now known as the Center for
Climate and Energy Solutions (C2ES)), World Meteorological Organization (WMO), and
the International Panel on Climate Change (IPCC) (State of California 2011) provide
information about climate change and its impact on the world through several
approaches.
2
These approaches include: (1) nonpartisan opinions about climate change (C2ES
2014); (2) the framework for global scale cooperation for meteorology and climate
(WMO 2014); and (3) assessing the current status of climate change (IPCC 2014).
Additionally, there are thousands of climate scientists throughout the world who are
studying the cause and effect relationship of climate change. More specifically, the
concern for this study is the effect on temperature due to climate change and ninety-seven
percent of these scientists indicate that humans are impacting the global climate change
(Anderegg et al. 2010, Doran et al. 2009, and Oreskes 2004).
The rise in global surface temperature anomaly (°C) is depicted in Figure 1 and is
based on the results from four scientific institutions: NASA Goddard Institute for Space
Studies (GISS), Met Office Hadley Centre/Climatic Research Unit (MOHC), NOAA
National Climatic Data Center (NCDC), and the Japanese Meteorological Agency (JMA).
An anomaly is the current climate variable’s departure from average conditions for a
particular place over a specific time period. Actual temperature observations from 1880
to the late 1930’s for all four institutions demonstrate a cooling trend with temperatures at
the greatest nearly -0.55°C below average temperatures. Additionally, the maximum
temperature anomaly occurred since the late 1970’s with an increase of nearly 0.65°C and
the observed temperatures are greater than the average temperature for all four
institutions since the late 1970’s. This steady increase in temperature is unforeseen since
1880 and therefore is evidence of a changing climate. The following subchapters offer
greater detail covering temperature trends at the global and local scale.
3
Figure 1: Global surface temperature anomaly (°C) for four worldwide scientific
institutions from 1880-2020. These anomalies show a dramatic increase in temperature
since 1980.
Source: National Aeronautic and Space Administration
(http://climate.nasa.gov/scientific-consensus)
1.1.1 Climate Change across the World
According to the IPCC’s Fourth Assessment Report (2007), global temperature is
expected to rise from 2.5°F to 10°F over the next 100 years with projected changes at
various locations: 1) tropical rainforests will be replaced with savannah in the eastern
Amazon; 2) millions in Africa will suffer from increased water stress; 3) decrease in
western North American mountain snowpack; 4) flash flooding increased in Europe; and
5) increased flooding along the Asian coastline. The positive or negative response to
climate change, such as the five listed above, is solely dependent upon the characteristics
of the individual species and ecosystems (Beaumont et al. 2011).
4
Global temperatures are increasing at a pace that does not seem to be slowing
down and is further clarified by NASA’s observation that “nine of the ten warmest years
have occurred since 1998” (NASA 2014b). This statement is visually explained in Figure
2 where the graph measures temperature anomaly from 1901-2013 in degree Fahrenheit
and shows that a fluctuation in negative and positive temperature anomalies occur until
the late 1970’s where only positive anomalies occur from this point forward. These
positive anomalies indicate that temperatures have been increasing since 1901. This
warming trend is further evident in the year 2012 as the ninth warmest global average
surface temperature since 1880 was recorded (NASA 2014b). A study by Hansen et al.
(2006), show global surface temperature has increased by approximately 0.2°C per year
within the last three decades which is the expected warming determined by the 2007
IPCC’s Fourth Assessment Report. Figure 3 shows that global surface temperature will
nearly quadruple with estimated temperatures of 1 to 1.5°F in 2020 to 4.5 to 7.5°F in
2099 (2007). Furthermore, scientific evidence shows that the probability is less than 5
percent that the increase in global surface temperature is caused by anything other than
anthropogenic climate change (greenhouse gas forcing) as compared to internal climate
variability (IPCC 2007).
5
Figure 2: Global surface temperature anomaly (°F) from 1901-2013. The bar graph
depicts actual temperature measurements with a positive (red) temperature anomaly or a
negative (blue) temperature anomaly. The linear trend line represents satellite
measurements in the lower troposphere and analyzed by two different groups: University
of Alabama-Huntsville (UAH) and Remote Sensing Systems (RSS).
Source: United States Environmental Protection Agency
(www.ncdc.noaa.gov/oa/ncdc.html)
6
Figure 3 Global average surface temperature (°C) projections for 2020-2029 and 2090-
2099.
Source: International Panel on Climate Change
(http://www.ipcc.ch/publications_and_data/ar4/wg1/en/spmsspm-projections-of.html)
1.1.2 Climate Change in the United States
According to the United States Environmental Protection Agency (2014), the
United States is reporting that the contiguous 48 states experienced a 0.14°F increase in
average surface temperature per decade since 1901. Figure 4 illustrates the fluctuation of
surface temperature since 1901 with the most distinguishable temperature change
occurring from the late 1990s to 2013 with a difference of more than 3°F. Also, the
average surface temperature increased from 0.31°F to 0.48°F per decade starting in the
late 1970s. Additionally, Karl et al. (2009, p. 9) states: “The winter months are
undergoing the greatest warming trend in the last 30 years.” An example from the report
is that they indicate that the Midwest (i.e., Illinois, Indiana, Iowa, Michigan, Minnesota,
Missouri, Ohio, and Wisconsin) and the Northern Great Plains (i.e., Montana, Nebraska,
7
North Dakota, South Dakota, and Wyoming) is a heavily impacted region with an
increase of 7°F during these months (2009).
Figure 4: U.S. temperature anomaly (°F) from 1901-2013 across the contiguous 48 states.
The bar graph depicts actual temperature measurements with a positive (red) temperature
anomaly or a negative (blue) temperature anomaly. The linear trend line represents
satellite measurements in the lower troposphere and analyzed by two different groups:
University of Alabama-Huntsville (UAH) and Remote Sensing Systems (RSS).
Source: United States Environmental Protection Agency
(www.ncdc.noaa.gov/oa/ncdc.html)
8
1.1.3 Climate Change in Southern California
Southern California, as referred to by the Southern California Association of
Governments (2009), is composed of six counties (i.e., Imperial, Los Angeles, Orange,
Riverside, San Bernardino, and Ventura) within a Mediterranean climate. Mediterranean
climate zones typically experience wet winters with relatively warm temperatures and dry
summers with hot temperatures (Marietta College 2014). While this climate zone’s high
temperatures and low precipitation is normal, indication of temperature rise and climate
change are found in Southern California.
The Office of Environmental Health Hazard Assessment (OEHHA) finds an
increase in annual average temperatures statewide and more specifically in Southern
California of 1.5°F per century beginning the year of 1895 (OEHHA 2013). Figure 5
clearly illustrates a temperature increase over the past 117 years with a sustained increase
in temperature of approximately 1°F since the 1970’s. It is important to note that the
statewide temperature data for this one reporting system are monthly average
temperatures acquired by the National Weather Service Cooperative Network (COOP)
observers and Parameter-elevation Regressions on Independent Slopes Model (PRISM)
data for 195 COOP stations throughout California (Western Regional Climate Center
2014).
Figure 6 illustrates that the departure from the average increases for the mean
temperature, minimum temperature, and maximum temperature starting in 1895 (Figure
6). OHEAA states that according to Figure 6 the fastest increase in temperature is
minimum temperature with a 1.99°F increase per 100 years; on the other hand, maximum
temperatures only increased at a rate of 1.01°F per 100 years (2013). More specifically,
9
the South Coast Region which includes the Los Angeles Basin and San Diego has
experienced a warming trend since 1895 (2013). Figure 7 clarifies the temperature trend
for the South Coast region where the annual departure values are derived from 1949 to
2005 averages. Overall, temperature is increasing in Southern California and this
temperature increase is an indication that climate change is a factor.
Figure 5: Annual average temperature trend for the State of California (1895-2012). The
bold line is the 11-year running mean for 195 COOP stations statewide.
Source: Office of Environmental Health Hazard Assessment
(http://www.wrcc.dri.edu/monitor/cal-mon/index.html)
10
Figure 6: Departure from average for mean temperature, minimum temperature, and
maximum temperature for the State of California (1895-2012). The bold line is 11-year
running mean and the thin line is the departure from the mean for 195 COOP stations
statewide.
Source: Office of Environmental Health Hazard Assessment
(http://www.wrcc.dri.edu/monitor/cal-mon/index.html)
11
Figure 7: Departure from average for mean temperature, minimum temperature, and
maximum temperature for the South Coast Region in the State of California. The bold
line is 11-year running mean and the thin line is the departure from the mean for a region
between Point Conception and the Mexico border.
Source: Office of Environmental Health Hazard Assessment
(http://www.wrcc.dri.edu/monitor/cal-mon/index.html)
1.2 Research Questions
According to recent scientific studies (Hansen et al. 2006; Intergovernmental
Panel on Climate Change 2007; Karl et al. 2009; Office of Environmental Health Hazard
Assessment 2013; and United States Environmental Protection Agency 2014), a changing
climate at various scales (global to local) is easily confirmed. Therefore, records of
historical climate data can provide evidence of historical temperature trends in an area.
The goal of this study is to analyze and interpret historical temperature data from 1931 to
2010 in Los Angeles County. Specifically, this study attempted to answer the research
questions below:
12
1. What are the roles of daily, monthly, and yearly temperature to interpret the historical
temperature trends in Los Angeles County?
2. How have mean temperature and extreme temperature thresholds changed and what
are the characteristics of these trends (changes) across Los Angeles County?
13
CHAPTER TWO: LITERATURE REVIEW
The following are scientific studies describing the impact of climate change on
temperature changes, study of extreme temperature thresholds, and the urban heat island
effect across the globe.
2.1 Climate Change and its Impact on Temperature Changes
Climate change affects global temperatures and researchers are studying and
analyzing its effects to provide scientific models and evidence that our planet is warming
at an unforeseen rate. Schlesinger (2011) discusses how the amount of incoming solar
radiation has increased due to human-induced greenhouse gases. According to
Schlesinger (2011), air temperature is expected to increase between 2°C to 4.5°C because
of the greenhouse-gas impact. Implications of this warming trend are explained by
Schlesinger when he states that the ocean’s warming temperature, slower than the Earth’s
atmosphere, increases the rate of evaporation, and in turn increases the amount of water
vapor in the atmosphere and an increase in the absorption of incoming solar radiation
(Schlesinger 2011). The effects of climate change and a warming atmosphere is felt in
the United States and a regional study of the United States is described below.
A temperature analysis of the western United States uses daily temperature and
precipitation data from 1950 to 2005 to monitor temperature changes over six different
regions (Booth et al. 2012). The main finding of this study is that climate change is
impacting the western United States and historical temperature data verifies this trend.
Additionally, Booth et al. (2012) discover that the California-Nevada region is
undergoing a trend favoring increasing daily minimum temperatures and a decreasing
14
number of frost days. Another important finding during this study is the lack of any
significant trends for maximum temperature in the California-Nevada region (Booth et al.
2012). Even though there is no apparent trend for maximum temperature, there is a
defined warming trend for northern and southern California with coastal regions of the
state experiencing a cooling trend during the time series (Booth et al. 2012). Further
evidence that climate change is occurring in California is described by the following
studies.
Cordero et al. (2011) analyze climatic data from 1918 to 2006 for the State of
California. According to their study, minimum and maximum temperatures are
increasing significantly across the entire State of California. As a result, minimum
temperatures increased by 0.17°C per decade while maximum temperatures increased by
0.07°C per decade (Cordero et al. 2011). Also, the study finds that Southern California is
undergoing the largest warming trend in California with a greater warming trend
occurring with maximum temperature rather than minimum temperature in Southern
California (Cordero et al. 2011).
Projection studies provide further evidence that a changing climate, more
specifically a warming climate, is occurring in California (Cayan et al. 2008). Two
different climate models (Parallel Climate Model and NOAA Geophysical Fluid
Dynamics Laboratory CM2.1) are tested to identify the type of temperature trend taking
place in California. The models produced results that include an increase in temperature
across California during the twenty-first century with projections ranging from 1.7°C to
5.8°C from 2000 to 2100 (Cayan et al. 2008). Gregory Bohr (2009) analyzes daily
maximum and minimum temperature data in California for 44 rural and 46 urban sites
15
from 1950 to 2005. The overall trend is warming temperatures statewide with an increase
in daily minimum and maximum temperatures (Bohr 2009). Also, California is
undergoing the largest increase in temperature from daily minimum temperatures (Bohr
2009). Also, Bohr’s findings show that largest temperature difference occurs with
warmer daily temperature minimums and cooler daily temperature maximums compared
to hot summer months maximum daily temperature and winter’s cold minimum daily
temperature. The effects of a changing climate are analyzed at metropolitan areas across
the United States and their results are described by the following.
Vimal Mishra and Dennis Lettenmaier (2011) analyze climate data for 100 of the
largest cities in the continental United States spanning 1950 to 2009. The author’s results
include a significant decreasing trend in heating degree days across the United States
with approximately 50 percent of the metropolitan areas experiencing this decline
(Mishra et al. 2011). Another important find is a statistically significant increase in warm
nights with 6.5 percent of the metropolitan areas indicating a warming trend and a
statistically significant declining trend of cool nights is predominant for the same
metropolitan areas in this study (Mishra et al. 2011). Overall, a strong warming trend is
occurring across the United States and is a clear indication that our climate is changing
and temperatures are increasing.
Taha (1997) explains that the human influence on climate change (anthropogenic
climate change) has the potential to affect near-surface temperatures in urban areas. The
findings show that anthropogenic temperature fluxes are the largest during the winter
months at cold climate metropolitan regions (Taha 1997). Also, day and night
temperatures are expected to rise between 2°C to 3°C in an urban region due to
16
anthropogenic temperature fluxes (1997). Overall, Taha states that an increase in
temperature resulting from anthropogenic forcing plays a role in urbanized areas, but this
temperature increase is “negligible in residential and commercial areas” (1997, p. 102).
Global surface temperature change is analyzed from 1870 to 1990 and 1998 to
2008 to determine the role of anthropogenic forcing on global temperature for this time
period (Kaufmann et al. 2006; Kaufmann et al. 2011). These two studies execute a
climate model that incorporates three equations using variables such as global surface
temperature, CO
2
, and CH
4
(Kaufmann et al. 2006). Kaufmann et al. (2006) find that the
global surface temperature increase is statistically significant from 1870 to 1990, and this
global surface temperature increase is most likely associated with greenhouse gases,
anthropogenic sulfur emissions, and solar activity. On the other hand, Kaufmann et al.
(2011) find that intensity of warming global surface temperature declines compared to
their previous study. This changing global temperature trend is related to an increase in
anthropogenic sulfur emissions which reduces the forcing effect of solar radiation
(Kaufmann et al. 2011). These two studies show that anthropogenic forcing can affect
global surface temperature in various ways and Kaufmann et al. state that “anthropogenic
factors have well known warming and cooling effects” (2011, p. 11792).
2.2 Threshold Temperature Analysis
In addition to maximum, minimum, and mean temperatures, scientists also
examine temperature thresholds as an indicator of temporal variability of temperature
(Meehl et al. 2004; Ruddell et al. 2013). Recently, Ruddell et al. (2013) studied the
temporal variability of temperature in Phoenix and Gila Bend, Arizona. Ruddell and his
17
colleagues use three daily temperature threshold variables (i.e., frost day, misery day, and
extreme heat event) to measure temporal variability for multiple time-series in Phoenix
and Gila Bend. Ruddell et al. define frost days as any day with a minimum temperature
of less than 32°F, misery days is any with a maximum daily temperature value greater
than or equal to 110°F (Ruddell et al. 2013). The last temperature threshold variable is
an extreme heat event and three criteria are required to classify an event as extreme heat
event with T1 and T2 defined as the 97.5 percentile of the normal conditions and the 81
percentile of the normal conditions, respectively (Meehl et al. 2004; Ruddell et al. 2013).
The three criteria include: (1) three consecutive days of daily maximum temperature
above T1; (2) entire period must have T1 below the average daily maximum temperature;
and (3) entire period must have T2 below daily maximum temperature (2004, p. 995;
2013, p. 205).
The authors approach to the results includes dividing each weather station’s
temperature data into ten year increments from 1900 to 2007. The results show that a
decreasing number of frost days and an increasing number of misery days, especially
between 1970 and 2007, are occurring at Phoenix. Also, the number of extreme heat
events increase greatly over the same time period. Hence, these trends provide evidence
that Phoenix is experiencing an enhanced warming trend from 1900 to 2007 (Ruddell et
al. 2013). In contrast, Gila Bend experienced only a slight, even relatively stable,
warming trend from 1900 to 2007 with a decrease in frost days but only a slight increase
in misery days (Ruddell et al. 2013). The threshold analysis results in changes of
temporal variability in temperature which can cause significant impacts on various
systems. Furthermore, this temporal study shows that an urbanized area like Phoenix is
18
experiencing the effects of climate change more than the rural area of Gila Bend,
Arizona.
Meehl and Tebaldi (2004) study extreme heat events using a global coupled
climate model known as the Parallel Climate Model (PCM). The two PCMs include a
four-member ensemble and a five-member ensemble which measures 20
th
century (1961
to 1990) and 21
st
century (2080 to 2099) climate variability and climate change for
extreme heat events at Chicago, Illinois and Paris, France (Meehl et al. 2004). The four-
member ensemble is a “model run four times from different initial states and the four
members are averaged together to reduce noise” and includes various forcing variables
(i.e., greenhouse gases, sulfate aerosols, ozone, volcanic aerosols, and solar variability) to
analyze heat wave events (2004, p. 994). The five-member ensemble model follows the
same averaging process as its predecessor (four-member ensemble model), but the model
is run five times and five members are averaged together to reduce noise. Also, this five-
member ensemble “assumes little in the way of policy intervention to mitigate
greenhouse gas emissions” (2004, p. 994). These two climate models are compared to
predict the characteristics of extreme heat events in these two locales based upon the
definition of a heat event (2004, p. 995). The four-member and five-member ensembles
results show that occurrences of heat waves and the duration of heat waves are predicted
to increase in the 21
st
century. The comparison between the four-member and five-
member ensemble models predicts a 25 percent increase in heat wave occurrences in
Chicago, Illinois and a 31 percent increase in heat wave occurrences in Paris, France.
Additionally, the duration of heat waves is predicted to increase at both Chicago and
Paris in the future. The five-member ensemble climate model expects the duration of
19
heat waves in Chicago, Illinois to be 8.47 to 9.24 days compared to the four-member
ensemble climate model that shows the current duration trend at 5.39 to 8.85 days. In
comparison, the five-member ensemble model predicts the duration of a heat wave to be
11.81 to 17.04 days compared to 8.33 to 12.69 days by the four-member ensemble model
at Paris, France.
2.3 Urban Heat Island Effect
Luke Howard (1833) is the first to document the impact of the urban heat island
on surface temperature and his study discussed temperature changes in London, England.
His study clarifies the urban heat island effect by finding that mean surface temperature
increases by 2°F within the urban area of London (Howard 1833). Another important
finding is the largest fluctuations in mean temperature occur during the winter months
and these large fluctuations are directly related to warm city nights where a difference
can be up to 3.7°F. Since Luke Howard, other scientists analyzed the relationship
between higher temperatures and urbanized areas, and specific studies results are detailed
by the following (Oke 1982; Camilloni et al. 1997; Taha 1997; Goodridge 1992).
Oke (1982) digs into the causes of the heat island effect at various levels of
atmosphere (i.e., urban canopy layer and urban boundary layer). His results are discussed
in Table 2 (1982, p. 17) with some key results explaining possible causes of the urban
heat island at various atmospheric levels: (1) increased absorption of short-wave radiation
at the canopy and boundary layer; (2) influence of anthropogenic heat sources at the
canopy and boundary layer; (3) decreased evapotranspiration at the canopy layer; (4)
increased incoming long-wave radiation and decreased long-wave radiation loss at the
20
canopy layer; (5) sensible heat storage increasing and total turbulent heat transport
decreasing at the canopy layer; and (6) increase of the sensible heat input-entrainment
above and below the boundary layer. Oke summarizes that the effect of the urban canopy
and boundary layer on the urban heat effect is not the same because the canopy layer is
related to “site character” and the boundary layer is impacted by the advection of warmer
air from above and “internal radiative effects” (1982, p. 21).
Camilloni et al. (1997) study the impact of the urban heat island and temperature
trends across Argentina, Australia, and the United States for closely located urban and
rural weather stations. Specifically, Camilloni et al. (1997) analyze 31 urban/rural pairs
for the yearly mean temperature difference °C (urban minus rural) and discover that the
urban regions year-to-year variability is less significant than its rural counterpart
(Camilloni et al. 1997). Camilloni et al. (1997) report a statistically significant cooling
trend of -0.04°C per year is discovered from 1925 to 1946 at the San Bernardino weather
station and a statistically significant warming trend of 0.03°C per year at the same station
from 1946 to 1968. Overall, the United States urban/rural paired stations experience a
warming trend before 1930 and a cooling trend after 1970 during these two time periods
(Camilloni et al. 1997).
Another urban heat island study involves modeling albedo, evapotranspiration,
and anthropogenic heating to discover the effect of urbanization on temperature changes
at various global cities, including Los Angeles, California (Taha 1997). The results for
Los Angeles include an albedo of 0.20 at the center of the city and an albedo difference
of 0.09 between the rural areas surrounding Los Angeles. Albedo is a key factor in
surface temperature in Los Angeles because an increase in albedo of 0.13 can decrease
21
temperatures between 2°C and 4°C (Taha 1997). Moreover, evapotranspiration plays
such an important role in an urban area that “evapotranspiration can create ‘oases’ that
are 2 to 8°C cooler than their surroundings” (1997, p. 101). The effect of anthropogenic
heating is examined in Los Angeles and Taha (1997) finds that the anthropogenic heating
measurement is 21 Wm
-2
and the net wave radiation is 108 Wm
-2
. As a comparison, the
maximum anthropogenic heating measurement is 159 Wm
-2
in Manhattan, New York and
the minimum measurement is 16 Wm
-2
in St. Louis, Missouri; the maximum net wave
radiation measurement is Los Angeles at 108 Wm
-2
and the minimum measurement is 18
Wm
-2
at Fairbanks, Alaska. Furthermore, the inclusion of anthropogenic heating at any
large urban area can increase surface temperature between 2°C and 3°C (Taha 1997).
Goodridge (1992) investigates the urban impact on long-term temperature trends
at 112 weather stations over an 80 year time period (1910 to 1989) using monthly mean
temperature measurements. The study reveals a warming trend across the entire State of
California with an increase in combined average annual temperature of 0.014°F per year
for all 112 weather stations (Goodridge 1992). Also, this warming trend is statistically
significant with an R
2
equal to 0.15 and provides further evidence of a warming trend
from 1910 to 1989. Additionally, according to Figure 2 (1992, p. 2) stations located
within Los Angeles County demonstrate an increase of 0.2°F to 0.4°F per year, and an
average annual temperature increase between 0.032°F (World Weather Record dataset)
and 0.014°F (World Weather Record and Historical Climatological Network datasets) at
urban locations across the state (Goodridge 1992).
22
CHAPTER THREE: DATA SOURCES AND METHODS
The existence of temperature records in Los Angeles County since the late 1870s
describes how climate has changed over the last 140 years in the county. Also, the daily,
monthly, and yearly historical surface temperature data is compared to recognize
temperature and extreme temperature threshold trends occurring throughout Los Angeles
County. This chapter provides a description of the study area in conjunction with the
necessary data and analytic methods.
3.1 Description of Study Area
The area for this investigation is Los Angeles County, located in the southwest of
California (Figure 8). It sustains a moderate climate with an average temperature in the
coldest month of December of 48.3° F and a temperature of 84.8°F in the warmest month
of August (rssWeather 2014). Notably, Los Angeles County is the most populous county
in the United States with a total population of 9,818,605 and a population density of
2,419 per square mile in 2010 (U.S. Census Bureau 2014b). Additionally, the estimated
population at Los Angeles County on July 1
st
, 2013 is 10.0 million and is approximately
4.7 more people than the second largest county of Cook County, Illinois (U.S. Census
Bureau 2014a and b). This current estimate from 2013 is an approximate increase of
more than 9.9 million people since 1900 (total population: 170,298) in Los Angeles
County (U.S. Census Bureau 1995). Moreover, the City of Los Angeles, the second most
populous metropolitan area in the United States with a total population of 3,792,621 in
2010, is located within Los Angeles County (National League of Cities 2013). The
current population and extreme changes in population over the last century is revealing a
23
changing climate. Hence, a projected increase in annual-mean surface temperature of
approximately 3°F to 5°F and an increase in extreme hot days of approximately zero to
fifty-five is expected to occur in Los Angeles County by the mid-21
st
century (Hall et al.
2012).
24
Figure 8: Los Angeles County and its location within the State of California. The grey
areas indicate 88 cities and metropolitan areas in Los Angeles County.
25
Also unique to Los Angeles County is its geographic character. The county is
surrounded by high elevation mountains, low-lying valleys, dry deserts, and miles of
Pacific coastline. More specifically, there is a total of 4,084 square miles of land area
with 1,875 square miles of mountains and 75 miles of coastline within this total land area
(County of Los Angeles 2014). The lowest point in Los Angeles County is in
Wilmington with an elevation of nine feet below sea level and the highest point is in
Mount San Antonio in the San Gabriel Mountain range with an elevation of 10,080 above
sea level (County of Los Angeles 2014). A portion of the Mojave Desert lies in the
northeastern portion of Los Angeles County. The Mojave desert is located between the
Great Basin Desert and the Sonoran and is also known as a “high desert” because its
elevation extent is greater than 2,000 feet above sea level (U.S. Department of the
Interior 2014; Michaelson 2009) Also, the elevation of the Mojave Desert influences its
average minimum and maximum daily temperatures during the winter and summer
months (Michaelson 2009).
26
3.2 Data
The following subchapters describe where daily and monthly temperature data are
acquired and the type of temperature variables available from each temperature dataset.
These datasets are the primary source for determining the trend in temperature and
extreme temperature thresholds across Los Angeles County.
3.2.1 Historical Surface Temperature
The study utilizes the wide-range and easy accessibility of historical daily and
monthly temperature data for Los Angeles County via the World Wide Web. The daily
surface temperature dataset is accessed through the Western Regional Climate Center
(WRCC) domain. Additionally, the monthly surface temperature dataset is obtained at
the National Oceanic and Atmospheric Administration (NOAA) and the State of
California Department of Water Resources (DWR) domain. The following describes in
greater detail the acquisition and the use of the daily and monthly surface temperature
datasets.
3.2.1.1 Daily Surface Temperature
Daily surface temperatures across Los Angeles County are acquired from the
WRCC website, and this dataset includes daily minimum temperature, daily maximum
temperature, and daily mean temperature for all WRCC stations in California. To
recognize a trend in extreme temperature thresholds, an 80 year time period is chosen
because the most complete daily temperature data spans from 1931 to 2010. The
selection results in six weather stations (Fairmont, Los Angeles, Palmdale, Pasadena,
27
Sandberg, and UCLA) that contain the required temperature data. These station’s
historical temperature data provide the required information to analyze the decadal trends
of frost days, misery days, and heat wave events.
3.2.1.2 Monthly Surface Temperature
There are hundreds of available weather stations in Los Angeles County from
NOAA’s NCDC database. The database offers monthly minimum temperature, monthly
maximum temperature, and monthly mean temperature spanning from 1931 to 2012.
Additionally, the State of California DWR is a contributor to monthly surface
temperature which provides monthly mean temperature data for the Los Angeles Civic
Center from 1878 to 2004. These two contributors provide monthly temperature data for
a total of 106 weather stations across Los Angeles County (Figure 9).
28
Figure 9: Distribution of 106 weather stations containing historical monthly surface
temperature across Los Angeles County, California. Each blue circle represents a single
station that currently or previously measured daily temperature. The grey areas indicate
88 cities and metropolitan areas in Los Angeles County.
29
3.3 Selection of Weather Stations
3.3.1 Weather Stations for Daily Surface Temperature
The methods for selecting the required stations begins by establishing the criteria
that each station must contain approximately 80 years of daily surface temperature data
from the WRCC temperature database. The manual selection discovers a total of six
weather stations that meet these criteria (Table 1; Figure 10). Table 1 is described briefly
by the following: 1) four stations contain 80 years of daily temperature data (i.e.,
Fairmont, Los Angeles, Palmdale, and Pasadena) and 2) two stations contain 78 years of
daily temperature data (i.e., Sandberg and UCLA). Also, the data completeness is at a
minimum in Sandberg with a percentage of 74.1 and at a maximum in Los Angeles with
data completeness of 99.9 percent (See details in Table 1). Additionally, three stations
(Fairmont, Sandberg, and UCLA) data completeness range between 74.1 percent and
76.0 percent, and the last three stations (Los Angeles, Palmdale, and Pasadena) data
completeness range between 99.3 percent and 99.9 percent.
Table 1: Operating time period for the six weather stations containing daily temperature
data between 1931 and 2010.
Weather
Station
First Month of
Collection
First Year of
Collection
Last Month of
Collection
Last Year of
Collection
Data
Completeness
(%)
Fairmont January 1931 August 2010 74.4
Los Angeles January 1931 December 2010 99.9
Palmdale April 1931 December 2010 96.5
Pasadena January 1931 December 2010 99.3
Sandberg January 1933 December 2010 74.1
UCLA January 1933 December 2010 76.0
30
Figure 10: Spatial location of six weather stations with 80 years or more of historical
daily temperatures. Each colored circle, as denoted in the legend to its respective station,
represents a single station that currently or previously measured daily temperature.
31
3.3.1.1 Characteristics of Daily Temperature Data at Six Stations
Table 2 describes the characteristics of the six weather stations that contain
historical daily temperature measurements that operated for more than 80 years within the
boundary of Los Angeles County. The stations listed in Table 2 are still in operation with
the earliest recorded measurements from the early 1930s. While these six stations started
operating over 80 years ago, the recorded measurements do not span the same length of
time. For instance, Fairmont, Los Angeles, Palmdale, and Pasadena start recording daily
temperature data in 1931; Conversely, Sandberg and UCLA started recording daily
temperature data in 1948.
The combination of Table 2 and Figure 10 show the spatial distribution of the six
weather stations across Los Angeles County. Three of the stations are located in northern
Los Angeles County (Fairmont, Palmdale, and Sandberg) and three of the stations are
located in southern Los Angeles County (Fairmont, Los Angeles, and UCLA). Figure 11
shows (assisted by aerial imagery) that Los Angeles, Pasadena, and UCLA are located in
an urban area while Fairmont and Palmdale are located in fairly rural, desert regions of
the county. Moreover, Sandberg is spatially located in the mountainous region of the San
Gabriel Mountain Range. Also, there is a large variance in elevation for all six weather
stations (Table 2). For example, there are weather stations located in low lying regions
with their average elevation listed by the following: (1) Los Angeles at 275 feet; (2)
Pasadena at 863 feet; and (3) UCLA at 433 feet. The other end of the spectrum includes
weather stations that are located in high elevation regions and their average elevation is
as follows: (1) Fairmont at 3,060 feet; (2) Palmdale at 2,628 feet; and (3) Sandberg at
4,517 feet.
32
Figure 11 visualizes how weather stations have maintained the same weather
station, but their latitude and longitude location have changed over time. For instance,
Los Angeles changed latitude and longitude location three times from 1931 to 2010 with
a greatest distance between station locations at more than 6,000 meters. Also, Sandberg
moved to three different latitude and longitude locations with the greatest distance at only
25 meters. Furthermore, Figure 11 visually describes the type of land use (i.e., desert,
mountain, or urban) at each weather station location. The results show that the Los
Angeles, Pasadena, and UCLA weather stations are located in urban areas. On the other
hand, the Fairmont and Palmdale weather stations are located in desert regions, and the
Sandberg weather station is located in a desolate, mountainous area.
33
Table 2: Characteristics of the six weather stations containing daily temperature data.
Weather
Stations
Operational Period Latitude Longitude
Elevation
(feet)
Land use
Fairmont
01/01/1931 -05/18/1999
05/18/1999-Present
34°42’00”
34°42’15”
-118°26’00”
-118°25’39”
3,060
3,060
Desert
Los Angeles
(Civic Center,
USC, & WB
City)
01/01/1931-12/31/1939
01/01/1940-07/13/1964
07/13/1964-11/21/1985
07/13/1964-07/31/1964
11/21/1985-06/24/1999
06/24/1999-07/07/2007
07/07/2007-Present
34°03’00”
34°03’04”
34°03’04”
34°03’04”
34°03’04”
34°03’04”
34°01’18”
-118°15’00”
-118°14’00”
-118°14’00”
-118°14’00”
-118°14’00”
-118°14’07”
-118°17’29”
361
312
270
312
270
230
171
Urban
Palmdale
04/01/1931-06/30/1948
07/01/1948-03/01/1952
03/01/1952-12/01/1962
12/01/1962-01/01-1982
01/01/1982-11/01/1993
11/01/1993-Present
34°35’00”
34°34’00”
34°35’00”
34°35’00”
34°35’00”
34°35’16”
-118°07’00”
-118°07’00”
-118°07’00”
-118°06’00”
-118°06’00”
-118°05’39”
2,661
2,651
2,661
2,602
2,596
2,596
Desert
Pasadena
01/01/1931 -06/01/1952
06/01/1952-09/12/2000
09/12/2000-02/11/2010
02/11/2010-Present
34°09’00”
34°09’00”
34°08’54”
34°08’53”
-118°09’00”
-118°09’00”
-118°08’41”
-118°08’40”
860
864
864
864
Urban
Sandberg
WSMO
01/01/1933-01/01/1982
01/01/1982-04/01/1996
04/01/1996-09/12/2000
09/12/2000-Present
34°45’00”
34°45’00”
34°44’37”
34°44’37”
-118°44’00”
-118°44’00”
-118°43’28”
-118°43’27”
4,524
4,517
4,517
4,510
Mountain
UCLA
01/01/1933-05/09/1957
05/09/1957-09/12/2000
09/12/2000-Present
34°04’00”
34°04’00”
34°04’11”
-118°27’00”
-118°27’00”
-118°26’34”
440
430
430
Urban
34
Figure 11: Large-scale aerial imagery representation of six weather stations across Los
Angeles County. The six images, located in the top to rows, show the type of land cover
and the distance between each station’s measuring locales. The image on the bottom row
is a small-scale spatial representation of the six weather stations with eighty years or
more of daily temperature measurements.
3.3.2 Weather Stations for Monthly and Yearly Surface Temperature
A fitness to use method is used to locate widely distributed weather stations
because multiple weather stations have relocated without changing the station’s name.
Therefore, a multitude of processes are executed using Esri ArcMap 10.2.2 and Microsoft
Excel 2010 to discover the final selection of weather stations.
A point location for each weather station is identified by executing the ArcMap
tool Display X & Y Coordinates using each station’s unique identifying coordinates
35
within the NCDC and DWR monthly temperature dataset. After running this tool, a point
feature is created and displayed in ArcMap depicting their latitude and longitude
locations. An export of these point features creates two new shapefiles with one
shapefile containing the NCDC monthly temperature data and the second containing
DWR monthly temperature data. A consolidation of these two shapefiles is obtained by
executing the ArcMap tool Merge, which in turn creates a new, single shapefile with all
the monthly temperature data within its attribute table. Additionally, the Dissolve tool is
used to create a single row of information for each station by their respective latitude and
longitude, and the end process is a single table containing 106 rows, or 106 weather
stations.
After analyzing the new shapefile’s attribute table, the dataset contains weather
stations that have the same station name but contain different latitude and longitude
coordinates. The question becomes: “Should each station be considered as a separate
entity or create one central location (centroid) for the stations that have the same name?”
This question is answered by executing the Near ArcMap tool because it calculates the
distance from one station to the (next) nearest station. After examining the results, it is
determined to create a centroid for the weather stations that share the same station name
within 1 kilometer (km) of its (next) nearest weather station.
The creation of the centroid for the (next) nearest station is a complicated process.
The first objective is to create a 1 km buffer using the Buffer tool in the ArcMap Analysis
toolbox for all one hundred six weather stations. Once the Buffer tool is executed, there
is a manual selection of the weather stations within a 1 km radius and the operation of the
Mean Center tool within the ArcMap Spatial Statistics Toolbox. This Mean Center tool
36
is a centroid process that includes selecting all the stations to be included as one, central
location, which in return creates a new shapefile. After performing the same procedure
for all the LBS within the 1 km buffer, a total of 22 new shapefiles are created with a new
centroid or mean center. Now that all the centroids are created, a new CSV spreadsheet
is created and imported into ArcMap. Next, the Display X & Y Coordinates tool is
implemented to plot sixty-six weather stations in Los Angeles County.
The final selection of the weather stations is straight-forward and its purpose is to
analyze historical surface temperature data at various time periods across Los Angeles
County. The objective is to analyze complete data (temperature values for all 12 months
of each collected year), and to analyze consecutive 20 year and 60 year data for as many
stations as possible. While these objectives are reasonable, temperature data is not
always complete for each station. Since inconsistencies exist, subjective reasoning
should be applied for the final selection of the weather station. The final selection
includes 21 weather stations with monthly surface temperature measurements stretching
from 1931 to 2010 (Table 3). All 21 stations (Figure 12) provide the necessary spatial
and temporal material to assist in locating and analyzing surface temperature trends in
Los Angeles County.
37
Table 3: Operating time period for the 21 weather stations containing monthly surface
temperature data between 1931 and 2010.
Weather Station
First Month of
Collection
First Year of
Collection
Last Month of
Collection
Last Year of
Collection
Claremont Pomona
College
January 1931 December 1980
Culver City January 1935 June 1967
Fairmont January 1931 October 2012
LAX August 1944 November 2012
Llano January 1931 April 1945
Long Beach
Aquarium
January 1931 November 1969
Los Angeles Terminal
Annex
November 1940 December 1952
North Hollywood January 1936 March 1950
OPIDS January 1933 May 1958
Palmdale April 1931 June 1948
Pasadena January 1931 November 2012
Pomona Fairplex January 1931 December 1969
San Fernando January 1931 June 1961
San Pedro January 1931 August 1964
Sierra Madre Henszey January 1931 June 1958
Table Mountain January 1931 August 1961
Torrance Airport January 1932 August 1955
UCLA January 1933 November 2012
USC January 1878 August 2012
Valyermo FD February 1938 November 1971
Woodland Hills
Pierce College
July 1949 November 2012
38
Figure 12: Selected 21 weather stations across Los Angeles County, California for
monthly surface temperature. Each blue circle represents a single station that currently or
previously measured daily temperature. The grey areas indicate 88 cities and
metropolitan areas in Los Angeles County.
39
3.3.2.1 Weather Stations for Monthly Surface Temperature
The selection of the necessary weather stations is performed manually by
selecting stations from Table 1 and Table 3 that that meet the criteria that more than 40
years of monthly surface temperature data exists. After performing the manual selection,
a total of eight stations exist matching the predetermined criteria (Table 4; Figure 13).
The length of time varies for some of the stations and the following statistics state the
number of station(s) for each length of time, name of the station(s), and the length of time
with monthly temperature data: 1) three stations for 80 years (i.e., Fairmont, Palmdale,
and Pasadena); 2) two stations for 78 years (i.e., Sandberg and UCLA); 3) one station for
69 years (i.e., USC); 4) one station for 68 years (i.e., LAX); and 5) one station for 63
years (i.e., Woodland Hills Pierce College). Also, the eight stations data completeness
range between 72.9 percent (Sandberg) to 100.0 percent (LAX). Additionally, seven out
the eight weather stations (excluding Sandberg) contain more than 96.0 percent data
completeness (Table 4).
40
Table 4: Operating time period for the eight weather stations containing monthly surface
temperature data between 1931 and 2010.
Weather
Station
First Month of
Collection
First Year of
Collection
Last Month of
Collection
Last Year of
Collection
Data
Completeness
(%)
Fairmont January 1931 August 2010 96.0
LAX August 1944 December 2010 100.0
Palmdale April 1931 December 2010 98.2
Pasadena January 1931 December 2010 99.9
Sandberg January 1933 December 2010 72.9
UCLA January 1933 December 2010 99.6
USC January 1950 December 2010 99.6
Woodland
Hills Pierce
College
July 1949 December 2010 99.6
41
Figure 13: Spatial location of eight weather stations with approximately 80 years of
historical monthly temperatures. Each blue circle represents a single station that
currently or previously measured daily temperature. The grey areas indicate 88 cities and
metropolitan areas in Los Angeles County.
42
3.3.2.2 Weather Stations for Yearly Surface Temperature
Yearly surface temperature is analyzed for trends over two different time periods
using averaged monthly mean surface temperature data. The time periods include 1931
to 1950 and 1951 to 2010 to determine if specific trends exist at shorter (20 years) and
longer (60 years) time periods. The selection of the specific stations is completed by
manually selecting stations from Table 3 that fall within one or both time periods. After
each station is selected for its respective time period, a total of six stations span from
1951 to 2010 (Table 5) and 20 stations fall within the period from 1931 to 1950 (Table
6). Table 5 shows that the data completeness for all six stations range from 91.5 percent
(USC) to 100.0 percent (LAX and UCLA). Also, the range in data completeness for
Table 6 is 93.1 percent (Los Angeles Terminal Annex and OPIDS) to 100.0 percent
(Culver City, Pasadena, Pomona Fairplex, Table Mountain, and USC).
Table 5: Operating time period for the six weather stations containing yearly surface
temperature data between 1951 and 2010.
Weather
Station
First Month of
Collection
First Year of
Collection
Last Month of
Collection
Last Year of
Collection
Data
Completeness
(%)
Fairmont January 1951 August 2010 94.9
LAX January 1951 December
2010
100.0
Pasadena January 1951
December 2010
99.9
UCLA January 1951
December 2010
100.0
USC January 1951
December 2010
91.5
Woodland
Hills Pierce
College
January 1951
December 2010
99.9
43
Table 6: Operating time period for the 20 weather stations containing yearly surface
temperature data between 1931 and 1950.
Weather
Station
First Month of
Collection
First Year of
Collection
Last Month of
Collection
Last Year of
Collection
Data
Completeness
(%)
Claremont
Pomona
College
January 1931 December 1950 99.2
Culver City January 1935 December 1950 100.0
Fairmont January 1931 December 1950 99.9
LAX August 1944 December 1950 91.7
Llano January 1931 April 1945 99.4
Long Beach
Aquarium
January 1931 December 1950 97.9
Los Angeles
Terminal
Annex
November 1940 December 1950 93.1
North
Hollywood
January 1936 March 1950 95.3
OPIDS January 1933 December 1950 93.1
Palmdale April 1931 June 1948 94.2
Pasadena January 1931 December 1950 100.0
Pomona
Fairplex
January 1931 December 1950 100.0
San Fernando January 1931 October 1950 98.3
San Pedro January 1931 December 1950 96.7
Sierra Madre
Henszey
January 1931 December 1950 97.5
Table
Mountain
January 1931 December 1950 100.0
Torrance
Airport
January 1932 December 1950 96.9
UCLA January 1933 December 1950 98.1
USC January 1878 December 1950 100.0
Valyermo FD February 1938 December 1950 96.8
44
3.4 Analysis of the Climate Trend
3.4.1 Daily Temperature
The selection of weather stations using daily temperature data is complete and
analyzing extreme temperature threshold trends (i.e., frost day, misery day, and heat
wave events) for these six stations is the next step. The 80 years of daily maximum and
minimum temperature data are divided into ten year increments for each station (i.e.,
1931 to 1940, 1941 to 1950, 1951 to 1960, 1961 to 1970, 1971 to 1980, 1981 to 1990,
1991 to 2000, and 2001 to 2010) and analyzed for their respective extreme temperature
threshold (Ruddell et al. 2013). The statistical findings include the mean and total
number of frost and misery days, as well as threshold temperatures (T1 and T2),
frequency, intensity, and average duration of heat wave events. Furthermore, the linear
regression results for frost and misery days are analyzed for statistical significance using
the standard least squares regression model.
Frost days are defined as any day that has an observed minimum daily
temperature less than or equal to 32°F. Misery days are defined as any day that has an
observed maximum daily temperature greater than or equal to 90°F (Tamrazian et al.
2008). As for the heat wave event, T1 (97.5 percentile of the normal conditions) and T2
(81 percentile of the normal conditions) must match all three criteria derived from Meehl
et al. (2004, p. 995) and they are as follows: 1) three consecutive days of daily maximum
temperature above T1; 2) the entire study period must have T1 below the average daily
maximum temperature; and 3) entire study period must have T2 below daily maximum
temperature. The following define the remaining heat wave event variables: 1) frequency
45
is the number of heat wave events based on the determination of a heat wave event using
T1 and T2 daily maximum temperature; 2) intensity is the average maximum temperature
of the heat wave event; and 3) average duration is the average span of each heat wave
event (Ruddell et al. 2013).
3.4.2 Monthly Temperature
After the selection of each weather station is complete, the temperature trend for
all eight stations is graphed. This temperature trend models each station’s monthly mean
temperature change during their respective time period. These time periods span from
1931 to 2010, 1944 to 2010, 1949 to 2010, and 1950 to 2010. Lastly, the linear
regression results for monthly mean temperature are analyzed for statistical significance
using the standard least squares regression model.
3.4.3 Yearly Temperature
Once the selection of each weather station is complete, the yearly surface
temperature is calculated by averaging each year’s monthly mean surface temperature
data into a single or yearly temperature value for each time period. Now that the yearly
temperature measurements are calculated for each weather station, they are graphed for
two types of temperature trends during their respective time period (1931 to 1950 and
1951 to 2010). These temperature trends include: 1) yearly surface temperature and 2)
yearly summer surface temperature where summer is defined as including the months of
July thru September.
46
CHAPTER 4: RESULTS
The following chapter contains three approaches to analyze temperature and extreme
temperature threshold trends. The three approaches implement the use of historical daily,
monthly, and yearly surface temperature data for the period spanning from 1931 to 2010.
4.1 Daily Temperature and Its Trend
4.1.1 Frost and Misery Day Annual and Decadal Statistics
Table 7 and Figure 14 show that misery days are increasing and frost days are
decreasing at the Fairmont weather station spanning the entire 80 year time period.
Statistically, Fairmont’s misery days increases by 561 days during the full 70 years of
daily temperature data (1950 to 2010) and a total of 3,799 misery days during the 80 year
period (1931 to 2010). On the other hand, the number of frost days decreases by 71 days
during the 70 years of continuous daily temperature data (1950 to 2010) and a total of
1,689 frost days over the 80 year span (1931 to 2010). Additionally, the decade from
1941 to 1950 has 89 frost days and 154 frost days with an increase to 283 frost days and
596 misery days spanning from 1951 to 1960. This large increase in frost and misery
days is corresponding to the lack of daily temperature data spanning the decade from
1941 to 1950. The maximum and minimum number of frost days are 58 in 1987 and 3 in
1986, respectively. On the other hand, the maximum and minimum number of misery
days are 102 in 2003 and 37 in 2009.
47
Table 7: Total number and annual mean of frost days and misery days measured by
decade at the Fairmont weather station for a total of 80 years (1931 to 2010). Frost days
represent temperatures less than or equal to 32 degrees Fahrenheit and misery days are
greater than or equal to 90 degrees Fahrenheit.
Decade
Frost Days Misery Days
No. Mean No. Mean
Fairmont Weather Station
1931-1940 Incomplete Incomplete Incomplete Incomplete
1941-1950 89 8.9 154 15.4
1951-1960 283 28.3 596 59.6
1961-1970 281 28.1 579 57.9
1971-1980 271 27.1 539 53.9
1981-1990 333 33.3 590 59.0
1991-2000 220 22.0 626 62.6
2001-2010 212 21.2 715 71.5
Total 1,689 21.1 3,799 47.5
Figure 14: Annual number of frost days and misery days using daily temperature data at
the Fairmont weather station from 1948 to 2010. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 0.2207x + 49.266 *
y = -0.1381x + 33.712
0
20
40
60
80
100
1930 1940 1950 1960 1970 1980 1990 2000 2010
Number of Days
Year
Fairmont
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
48
Los Angeles undergoes an increase in misery days versus frost days with
more than 1,700 misery days than frost days over the 80 year period (Table 8; Figure 15).
Therefore, this period has a total of 10 frost days and 1,732 misery days. Also, the
number of misery days nearly double from 128 days (1941 to 150) to 229 days (1951 to
1960). The maximum and minimum number of frost days are 4 in 1949 and 0 for 74
different years of record, respectively. Conversely, the maximum and minimum number
of misery days are 47 in 1983 and 5 in 2001, respectively.
Table 8: Total number and annual mean of frost days and misery days measured by
decade at the Los Angeles weather station for a total of 80 years (1931 to 2010). Frost
days represent temperatures less than or equal to 32 degrees Fahrenheit and misery days
are greater than or equal to 90 degrees Fahrenheit.
Decade
Frost Days Misery Days
No. Mean No. Mean
Los Angeles Weather Station
1931-1940 2 0.2 150 15.0
1941-1950 4 0.4 128 12.8
1951-1960 1 0.1 229 22.9
1961-1970 0 0.0 207 20.7
1971-1980 3 0.3 239 23.9
1981-1990 0 0.0 290 29.0
1991-2000 0 0.0 277 27.7
2001-2010 0 0.0 212 21.2
Total 10 0.13 1,732 21.7
49
Figure 15: Annual number of frost days and misery days using daily temperature data at
the Los Angeles weather station from 1931 to 2010. The bold green line represents the
trend of misery events and the dashed green line represents the trend line for misery
events. The bold purple line represents the trend of frost events and the dashed purple
line represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
The Palmdale weather station experiences the most definitive contrast
between extreme temperature thresholds (Table 9; Figure 16). There are 4,234 frost days
and 8,345 misery days recorded in this 80 year period. The departure between these
events are 4,111 and is second only to Pasadena. The largest change in decadal trends
occurs during frost day events between the years of 1981 and 1990 (604) as well as 1991
and 2000 (395) with a difference of 209 frost days during this ten year period. The
maximum and minimum number of frost days are 86 in 1948 and 0 in 1931. Also, the
maximum and minimum misery days are 137 in 2003 and 41 in 1932, respectively. The
minimum number of frost and misery days are subjective due to the missing daily
temperature measurements from January 1
st
, 1931 until March 31
st
, 1931 and October 1
st
,
y = 0.1705x + 14.575 **
y = -0.0037x + 0.2801
0
10
20
30
40
50
1930 1940 1950 1960 1970 1980 1990 2000 2010
Number of Days
Year
Los Angeles
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
50
1931 until August 15
th
, 1932. Therefore, these inconsistencies are reflected in the trends
for frost and misery days within Figure 16.
Table 9: Total number and annual mean of frost days and misery days measured by
decade at the Palmdale weather station for a total of 80 years (1931 to 2010). Frost days
represent temperatures less than or equal to 32 degrees Fahrenheit and misery days are
greater than or equal to 90 degrees Fahrenheit.
Decade
Frost Days Misery Days
No. Mean No. Mean
Palmdale Weather Station
1931-1940 462 46.2 927 92.7
1941-1950 651 65.1 948 94.8
1951-1960 567 56.7 1,100 110.0
1961-1970 551 55.1 928 92.8
1971-1980 660 66.0 1,018 101.8
1981-1990 604 60.4 1,079 107.9
1991-2000 395 39.5 1,181 118.1
2001-2010 344 34.4 1,164 116.4
Total 4,234 52.9 8,345 104.3
Figure 16: Annual number of frost days and misery days using daily temperature data at
the Palmdale weather station from 1931 to 2010. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 0.3547x + 89.592 **
y = -0.22x + 62.054 **
0
20
40
60
80
100
120
140
1930 1940 1950 1960 1970 1980 1990 2000 2010
Number of Days
Year
Palmdale
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
51
Another weather station that experiences an increase in misery days and a
decrease in frost days over 80 years is Pasadena (Table 10; Figure 17), and the increase in
misery days over time is the most influential than any of the six weather stations
observed. Hence, Pasadena experiences the greatest range between frost and misery days
for 80 years with a difference of 4,445 days with a total of 142 frost days and 4,587
misery days. A noteworthy decadal trend occurs from 1931 to 1950 where the frost days
remain the same for 20 years (44 days) but the misery days decrease by 156 days (527 to
371 days). The maximum and minimum number of frost days are 16 in 1937 and 0 for
38 different years of record, respectively. Also, the maximum and minimum number of
misery days are 98 in 1990 and 24 in 1944, respectively.
Table 10: Total number and annual mean of frost days and misery days measured by
decade at the Pasadena weather station for a total of 80 years (1931 to 2010). Frost days
represent temperatures less than or equal to 32 degrees Fahrenheit and misery days are
greater than or equal to 90 degrees Fahrenheit.
Decade
Frost Days Misery Days
No. Mean No. Mean
Pasadena Weather Station
1931-1940 44 4.4 527 52.7
1941-1950 44 4.4 371 37.1
1951-1960 17 1.7 504 50.4
1961-1970 6 0.6 556 55.6
1971-1980 15 1.5 544 54.4
1981-1990 13 1.3 700 70.0
1991-2000 1 0.1 761 76.1
2001-2010 2 0.2 624 62.4
Total 142 1.8 4,587 57.3
52
Figure 17: Annual number of frost days and misery days using daily temperature data at
the Pasadena weather station from 1931 to 2010. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
Sandberg is the only weather station that demonstrates an increase in frost
days and a decrease in misery days over the 80 year period (Table 11; Figure 18).
Sandberg contains no daily temperature records from 1931 to 1940 and has missing daily
temperature data for a majority of the decade spanning from 1941 to 1950 (January 1
st
,
1941 to June 30
th
, 1948). This missing daily temperature data reveals biased results for
both frost and misery days spanning 1941 to 1950 where caution needs to be asserted
when analyzing this decadal trend. Therefore, the greatest decadal trend recognized is
from 1961 to 2000 because an increase of 77 frost days and a decrease of 365 misery
days exist for this time period. Another important finding is that while a trend in
decreasing misery days is discovered, the overall trend for misery reveals that there is a
larger number of total misery days (2,993) compared to frost days (1,256). Moreover, the
y = 0.3838x + 41.41 **
y = -0.0579x + 4.1788 **
0
20
40
60
80
100
1930 1940 1950 1960 1970 1980 1990 2000 2010
Number of Days
Year
Pasadena
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
53
maximum and minimum number of frost days are 49 in 1995 and 4 in 1963, respectively.
On the other hand, the maximum and minimum number of misery days are 90 in 1952
and 8 in 1995, respectively.
Table 11: Total number and annual mean of frost days and misery days measured by
decade at the Sandberg weather station for a total of 80 years (1931 to 2010). Frost days
represent temperatures less than or equal to 32 degrees Fahrenheit and misery days are
greater than or equal to 90 degrees Fahrenheit.
Decade
Frost Days Misery Days
No. Mean No. Mean
Sandberg Weather Station
1931-1940 Incomplete Incomplete Incomplete Incomplete
1941-1950 59 5.9 160 16.0
1951-1960 219 21.9 542 54.2
1961-1970 167 16.7 643 64.3
1971-1980 154 15.4 558 55.8
1981-1990 169 16.9 347 34.7
1991-2000 244 24.4 278 27.8
2001-2010 244 24.4 465 46.5
Total 1,256 15.7 2,993 37.4
Figure 18: Annual number of frost days and misery days using daily temperature data at
the Sandberg weather station from 1948 to 2010. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = -0.4369x + 69.354 **
y = 0.0944x + 15.215
0
20
40
60
80
100
1930 1940 1950 1960 1970 1980 1990 2000 2010
Number of Days
Year
Sandberg
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
54
The UCLA weather station is the last weather station that experiences an
increase in misery days and a decrease in frost days; however, the decrease in frost day
events is very subtle (Table 12; Figure 19). UCLA has no daily temperature records for
the decade spanning from 1931 to 1940 and a majority of the decade from 1941 to 1950.
The missing daily temperature data spans from January 1
st
, 1941 to June 30
th
, 1948 and
July 3
rd
, 1948 to July 11
th
, 1948. Once again, this missing daily temperature data
introduces biased results, urging caution in the results for this time period.
An important decadal trend discovered from 1961 to 2010 include no frost
days recorded for 50 years, and another decadal trend from 1941 to 2010 is the steady
increase in misery days for 70 years (21 to 101). The maximum and minimum number of
frost days are 2 for two years (1949 and 1957) and 0 for 61 different years of record,
respectively. Also, the maximum and minimum number of misery days are 21 for two
years (2008 and 2009) and 0 in 2001. The totals for the 80 year period are 4 frost days
and 536 misery days.
Table 12: Total number and annual mean of frost days and misery days measured by
decade at the UCLA weather station for a total of 80 years (1931 to 2010). Frost days
represent temperatures less than or equal to 32 degrees Fahrenheit and misery days are
greater than or equal to 90 degrees Fahrenheit.
Decade
Frost Days Misery Days
No. Mean No. Mean
UCLA Weather Station
1931-1940 Incomplete Incomplete Incomplete Incomplete
1941-1950 2 0.2 21 2.1
1951-1960 2 0.2 73 7.3
1961-1970 0 0.0 76 7.6
1971-1980 0 0.0 88 8.8
1981-1990 0 0.0 85 8.5
1991-2000 0 0.0 92 9.2
2001-2010 0 0.0 101 10.1
Total 4 0.05 536 6.7
55
Figure 19: Annual number of frost days and misery days using daily temperature data at
the UCLA weather station from 1931 to 2010. The bold green line represents the trend of
misery events and the dashed green line represents the trend line for misery events. The
bold purple line represents the trend of frost events and the dashed purple line represents
the trend line for frost events. A single asterisk (*) represents statistical significance at p-
value < 0.05 and a double asterisk (**) represents statistical significance a p-value < 0.01.
4.1.1.1 Frost and Misery Day Linear Regression Significance
The evaluation of each station’s frost and misery day linear regression line is
accomplished by determining the statistical significance of each extreme temperature
threshold. This evaluation process includes executing the standard least squares
regression with the independent variable (y) set as year and the dependent variable (x) set
as frost day or misery day. This regression analysis shows multiple statistical results
from the execution of the standard least square analysis (i.e., R
2
, adjusted R
2
, root mean
square error, sum of squares, F ratio, etc.); however, this analysis only investigates the
adjusted coefficient of determination (R
2
), slope coefficient, and the p-value < 0.05 (95
percent level) or p-value < 0.01 (99 percent level) statistics for the frost and misery day
linear regression line.
y = 0.0632x + 5.3469
y = -0.005x + 0.3131 *
0
5
10
15
20
25
1930 1940 1950 1960 1970 1980 1990 2000 2010
Number of Days
Year
UCLA
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
56
Table 13 and Table 14 categorize the statistical significance results for all six
weather stations for the entirety of their respective time period and their results are as
follows. Fairmont, Los Angeles, and Sandberg are not significant at either p-value level
during frost days, but misery days are statistically significant at the 95 percent level
(0.0209), the 99 percent level (0.0001), and the 99 percent level (0.0013), respectively.
Also, these three stations adjusted R
2
values are 0.0694, 0.1603, and 0.1438, respectively,
and the adjusted R
2
value indicates how efficiently the linear regression line represents
the misery day threshold data. On the other hand, UCLA is the only weather station that
experiences statistical significance only during frost days with a p-value at the 95 percent
level of 0.0405 and an adjusted R
2
value of 0.0517.
Palmdale and Pasadena experience statistically significant frost and misery days
at the 99 percent level. Palmdale’s frost days p-value is 0.0096 and adjusted R
2
is
0.0710, and the misery days p-value is 0.0001 and adjusted R
2
is 0.2438. Additionally,
Pasadena’s frost days p-value is 0.0001 and adjusted R
2
is 0.1769, and the misery days p-
value is 0.0001 and adjusted R
2
is 0.2958. Overall, all six stations show statistical
significance, either p-value < 0.05 or p-value < 0.01, during frost or misery days, and two
stations (Palmdale and Pasadena) demonstrate statistical significance at the 99 percent
level for both frost and misery days.
57
Table 13: Statistical significance characteristics for frost days at all six weather stations.
A single asterisk (*) represents statistical significance at p-value < 0.05 and a double
asterisk (**) represents statistical significance a p-value < 0.01.
Weather
Station
Data Time
Period
Adjusted R
2
Slope
Coefficient
p-value
Fairmont 1948-2010 0.0306 -0.1380 0.0903
Los Angeles 1931-2010 0.0099 -0.0037 0.1856
Palmdale 1931-2010 0.0710 -0.2199 0.0096 **
Pasadena 1931-2010 0.1769 -0.0579 0.0001 **
Sandberg 1948-2010 0.0167 0.0944 0.1569
UCLA 1948-2010 0.0517 -0.0049 0.0405 *
Table 14: Statistical significance characteristics for misery days at all six weather
stations. A single asterisk (*) represents statistical significance at p-value < 0.05 and a
double asterisk (**) represents statistical significance a p-value < 0.01.
Weather
Station
Data Time
Period
Adjusted R
2
Slope
Coefficient
p-value
Fairmont 1948-2010 0.0694 0.2207 0.0209 *
Los Angeles 1931-2010 0.1603 0.1705 0.0001 **
Palmdale 1931-2010 0.2438 0.3547 0.0001 **
Pasadena 1931-2010 0.2958 0.3838 0.0001 **
Sandberg 1948-2010 0.1438 -0.4369 0.0013 **
UCLA 1948-2010 0.0443 0.0632 0.0535
4.1.1.2 Spatial Discontinuity of Three Weather Stations and Their Trends
The results from the frost and misery day annual trends discover that a further
investigation into spatial discontinuity is required for Los Angeles, Palmdale, and
Sandberg. Since these stations are spatially dispersed the farthest apart for their
respective location (Figure 11), an investigation is required to determine the trend of frost
58
and misery days as well as the statistical significance of each linear regression line during
the weather station’s respective time period by using the standard least squares
regression. The parameter for the independent variable (y) is set as year and the
dependent variable (x) is set as frost day or misery day, and the listed results are the
adjusted coefficient of determination (R
2
), slope coefficient, and the p-value < 0.05 (95
percent level) or p-value < 0.01 (99 percent level). These three stations linear trend and
standard least squares regression results are as follows.
Figure 20, 21, 22, 23, 24, and 25 illustrate the occurring trend for frost and misery
days from 1931 to 1939, 1940 to 1964, 1964 to 1985, 1985 to 1999, 1999 to 2007, and
2007 to 2010 at the Los Angeles weather station, respectively. These six different
spatially located Los Angeles stations experience slight variability at each temperature
measuring instance based upon the linear regression model. A slight decrease in misery
days and increase in frost days occurs from 1931 to 1939, and a slight increase in misery
days and decrease in frost days occurs from 1940 to 1964. Also, the period from 1964 to
1985 experiences a modest increase in misery and frost days. Evidence suggests that a
warming trend occurred during these three time periods (1985 to 1999, 1999 to 2007, and
2007 to 2010) because of a linear increase in misery days and no recorded frost days
during these time periods. Additionally, the only extreme threshold linear regression line
that shows any statistical significance is misery days from 1964 to 1985 with an adjusted
R
2
of 0.2602 and p-value < 0.01 of 0.0054 (Table 15; Table 16).
59
Figure 20: Annual number of frost days and misery days using daily temperature data at
the Los Angeles weather station from 1931 to 1939. The bold green line represents the
trend of misery events and the dashed green line represents the trend line for misery
events. The bold purple line represents the trend of frost events and the dashed purple
line represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = -0.3167x + 16.583 y = 0.0667x - 0.1111
0
20
40
60
80
100
120
140
1931 1932 1933 1934 1935 1936 1937 1938 1939
Number of Days
Year
Los Angeles
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
60
Figure 21: Annual number of frost days and misery days using daily temperature data at
the Los Angeles weather station from 1940 to 1964. The bold green line represents the
trend of misery events and the dashed green line represents the trend line for misery
events. The bold purple line represents the trend of frost events and the dashed purple
line represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 0.5238x + 10.55 ** y = -0.0135x + 0.3903
0
20
40
60
80
100
120
140
1940 1942 1944 1946 1948 1950 1952 1954 1956 1958 1960 1962 1964
Number of Days
Year
Los Angeles
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
61
Figure 22: Annual number of frost days and misery days using daily temperature data at
the Los Angeles weather station from 1964 to 1985. The bold green line represents the
trend of misery events and the dashed green line represents the trend line for misery
events. The bold purple line represents the trend of frost events and the dashed purple
line represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 0.4743x + 20.182
y = 0.013x - 0.013
0
20
40
60
80
100
120
140
1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984
Number of Days
Year
Los Angeles
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
62
Figure 23: Annual number of frost days and misery days using daily temperature data at
the Los Angeles weather station from 1985 to 1999. The bold green line represents the
trend of misery events and the dashed green line represents the trend line for misery
events. The bold purple line represents the trend of frost events and the dashed purple
line represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 0.3x + 25.4
y = 0
0
20
40
60
80
100
120
140
1985 1987 1989 1991 1993 1995 1997 1999
Number of Days
Year
Los Angeles
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
63
Figure 24: Annual number of frost days and misery days using daily temperature data at
the Los Angeles weather station from 1999 to 2007. The bold green line represents the
trend of misery events and the dashed green line represents the trend line for misery
events. The bold purple line represents the trend of frost events and the dashed purple
line represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 1.9833x + 8.6389
y = 0
0
20
40
60
80
100
120
140
1999 2000 2001 2002 2003 2004 2005 2006 2007
Number of Days
Year
Los Angeles
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
64
Figure 25: Annual number of frost days and misery days using daily temperature data at
the Los Angeles weather station from 2007 to 2010. The bold green line represents the
trend of misery events and the dashed green line represents the trend line for misery
events. The bold purple line represents the trend of frost events and the dashed purple
line represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
Table 15: Statistical significance characteristics for frost days at Los Angeles. A single
asterisk (*) represents statistical significance at p-value < 0.05 and a double asterisk (**)
represents statistical significance a p-value < 0.01.
Weather
Station
Data Time
Period
Adjusted R2 Slope p-value
Los Angeles 1931-1939 -0.0571 0.0666 0.4758
Los Angeles 1940-1964 -0.0318 -0.0135 0.5945
Los Angeles 1964-1985 -0.0158 0.0129 0.4218
Los Angeles 1985-1999 Null 0.0000 Null
Los Angeles 1999-2007 Null 0.0000 Null
Los Angeles 2007-2010 Null 0.0000 Null
y = 1.2x + 22
y = 0
0
20
40
60
80
100
120
140
2007 2008 2009 2010
Number of Days
Year
Los Angeles
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
65
Table 16: Statistical significance characteristics for misery days at Los Angeles. A single
asterisk (*) represents statistical significance at p-value < 0.05 and a double asterisk (**)
represents statistical significance a p-value < 0.01.
Weather
Station
Data Time
Period
Adjusted R2 Slope p-value
Los Angeles 1931-1939 -0.1218 -0.3167 0.7275
Los Angeles 1940-1964 0.2602 0.5238 0.0054 **
Los Angeles 1964-1985 0.0479 0.4743 0.1669
Los Angeles 1985-1999 -0.0532 0.3000 0.5976
Los Angeles 1999-2007 0.2046 1.9833 0.1238
Los Angeles 2007-2010 -0.0500 1.2000 0.4523
Figure 26, 27, 28, 29, 30, and 31 illustrates that higher variability in frost and
misery days occur as compared to the Los Angeles weather station from 1931 to 1948,
1948 to 1952, 1952 to 1962, 1962 to 1982, 1982 to 1993, and 1993 to 2010. The linear
regression model identifies a noticeable frost day trend occurring from 1931 to 1933 with
an accelerated increase in frost days (71 days) and an accelerated decrease in frost days
occurs from 1948 to 1950 (47 days). On the other hand, an accelerated increase in misery
days (62 days) occurs from 1932 to 1933 and an accelerated decrease in misery days (69
days) occurs from 1960 to 1961.
Additionally, the linear regression model shows that the time periods from 1931
to 1948 and 1962 to 1982 experience an increasing trend in misery days and frost days.
Also, the two time periods (1952 to 1962 and 1993 to 2010) experience a decreasing
trend in misery days and frost days. The final two time periods spanning from 1948 to
1952 and 1982 to 1993 saw an increase in misery days and decrease in frost days which
suggests that a warming trend occurred during these two time periods. Also, the standard
66
least squares regression proves that the misery day linear regression line from 1931 to
1948 is statistically significant at p-value < 0.01 (99 percent level) and an adjusted R
2
value of 0.4153, and the misery day linear regression line is statistically significant at p-
value < 0.05 (95 percent level) and an adjusted R
2
value of 0.2937 from 1982 to 1993
(Table 17; Table 18).
Figure 26: Annual number of frost days and misery days using daily temperature data at
the Palmdale weather station from 1931 to 1948. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 0.9267x + 83.974
y = 2.8679x + 28.922 **
0
20
40
60
80
100
120
140
1931 1933 1935 1937 1939 1941 1943 1945 1947
Number of Days
Year
Palmdale
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
67
Figure 27: Annual number of frost days and misery days using daily temperature data at
the Palmdale weather station from 1948 to 1952. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 2.9x + 92.1
y = -5.6x + 78.2
0
20
40
60
80
100
120
140
1948 1949 1950 1951 1952
Number of Days
Year
Palmdale
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
68
Figure 28: Annual number of frost days and misery days using daily temperature data at
the Palmdale weather station from 1952 to 1962. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = -0.0545x + 106.51
y = -2.0273x + 66.618
0
20
40
60
80
100
120
140
1952 1954 1956 1958 1960 1962
Number o f Days
Year
Palmdale
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
69
Figure 29: Annual number of frost days and misery days using daily temperature data at
the Palmdale weather station from 1962 to 1982. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 0.3169x + 95.99
y = 0.2455x + 58.014
0
20
40
60
80
100
120
140
1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982
Number of Days
Year
Palmdale
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
70
Figure 30: Annual number of frost days and misery days using daily temperature data at
the Palmdale weather station from 1982 to 1993. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 1.5664x + 98.152
y = -2.0524x + 71.424 *
0
20
40
60
80
100
120
140
1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993
Numer of Days
Year
Palmdale
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
71
Figure 31: Annual number of frost days and misery days using daily temperature data at
the Palmdale weather station from 1993 to 2010. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
Table 17: Statistical significance characteristics for frost days at Palmdale. A single
asterisk (*) represents statistical significance at p-value < 0.05 and a double asterisk (**)
represents statistical significance a p-value < 0.01.
Weather
Station
Data Time
Period
Adjusted R2 Slope p-value
Palmdale 1931-1948 0.4153 2.8680 0.0023 **
Palmdale 1948-1952 0.0409 -5.6000 0.3585
Palmdale 1952-1962 0.1271 2.0272 0.1515
Palmdale 1962-1982 -0.0373 0.2454 0.6027
Palmdale 1982-1993 0.2937 -2.0524 0.0399 *
Palmdale 1993-2010 -0.0456 -0.2859 0.6177
y = -0.3127x + 120.03
y = -0.2859x + 38.66
0
20
40
60
80
100
120
140
1993 1995 1997 1999 2001 2003 2005 2007 2009
Number of Days
Year
Palmdale
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
72
Table 18: Statistical significance characteristics for misery days at Palmdale. A single
asterisk (*) represents statistical significance at p-value < 0.05 and a double asterisk (**)
represents statistical significance a p-value < 0.01.
Weather
Station
Data Time
Period
Adjusted R2 Slope p-value
Palmdale 1931-1948 0.0269 0.9267 0.2428
Palmdale 1948-1952 0.3910 2.9000 0.1553
Palmdale 1952-1962 -0.1110 -0.0545 0.9779
Palmdale 1962-1982 -0.0210 0.3169 0.4528
Palmdale 1982-1993 0.1148 1.5664 0.1503
Palmdale 1993-2010 -0.0441 -0.3127 0.6027
As compared to Palmdale, Sandberg shows high misery and frost day variability
from 1948 to 1981, 1982 to 1996, 1996 to 2000, and 2000 to 2010 (Figure 32; Figure 33;
Figure 34; Figure 35). The linear regression shows that an accelerated shift in frost days
occurs during two time periods: (1) accelerated increase (34 frost days) occurs from 1993
to 1994 and (2) accelerated decrease (34 frost days) occurs from 1960 to 1963. Also, two
time periods experience a shift in misery days: (1) accelerated increase (45 misery days)
occurs from 1986 to 1987 and (2) accelerated decrease (64 misery days) occurs from
1975 to 1981.
Specifically, the linear regression model identifies 1948 to 1981 as a time period
that experiences a decrease in misery and frost days. Also, an accelerated increase in
misery days and decrease in frost days occurs from 1996 to 2000 and 2000 to 2010 which
suggests that a warming trend is occurring at Sandberg. Conversely, an accelerated
decrease in misery days and an accelerated increase in frost days occurs from 1982 to
1996 and this threshold trend suggests that a cooling trend occurs at Sandberg.
73
Furthermore, the results from the standard least squares regression model (Table 19;
Table 20) identify only one statistically significant linear regression line as frost days
from 1982 to 1996 with a p-value of 0.0029 (significant at the 99 percent level) and an
adjusted R
2
value of 0.4691.
Figure 32: Annual number of frost days and misery days using daily temperature data at
the Sandberg weather station from 1948 to 1981. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = -0.2906x + 61.556
y = -0.1584x + 21.155
0
20
40
60
80
100
120
140
1948 1951 1954 1957 1960 1963 1966 1969 1972 1975 1978 1981
Number of Days
Year
Sandberg
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
74
Figure 33: Annual number of frost days and misery days using daily temperature data at
the Sandberg weather station from 1982 to 1996. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = -1.6536x + 42.229 y = 2.1429x + 4.2571 **
0
20
40
60
80
100
120
140
1982 1984 1986 1988 1990 1992 1994 1996
Number of Days
Year
Sandberg
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
75
Figure 34: Annual number of frost days and misery days using daily temperature data at
the Sandberg weather station from 1996 to 2000. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
y = 4.3x + 25.5 y = -0.9x + 21.5
0
20
40
60
80
100
120
140
1996 1997 1998 1999 2000
Number of Days
Year
Sandberg
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
76
Figure 35: Annual number of frost days and misery days using daily temperature data at
the Sandberg weather station from 2000 to 2010. The bold green line represents the trend
of misery events and the dashed green line represents the trend line for misery events.
The bold purple line represents the trend of frost events and the dashed purple line
represents the trend line for frost events. A single asterisk (*) represents statistical
significance at p-value < 0.05 and a double asterisk (**) represents statistical significance
a p-value < 0.01.
Table 19: Statistical significance characteristics for frost days at Sandberg. A single
asterisk (*) represents statistical significance at p-value < 0.05 and a double asterisk (**)
represents statistical significance a p-value < 0.01.
Weather
Station
Data Time
Period
Adjusted R2 Slope p-value
Sandberg 1948-1981 0.0103 -0.1584 0.2551
Sandberg 1982-1996 0.4691 2.1429 0.0029 **
Sandberg 1996-2000 -0.3039 -0.9000 0.8115
Sandberg 2000-2010 -0.1015 -0.2000 0.7851
y = 1.6273x + 35.509
y = -0.2x + 25.473
0
20
40
60
80
100
120
140
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Number of Days
Year
Sandberg
Misery Days Frost Days Linear (Misery Days) Linear (Frost Days)
77
Table 20: Statistical significance characteristics for misery days at Sandberg. A single
asterisk (*) represents statistical significance at p-value < 0.05 and a double asterisk (**)
represents statistical significance a p-value < 0.01.
Weather
Station
Data Time
Period
Adjusted R2 Slope p-value
Sandberg 1931-1981 -0.0095 -0.2906 0.4119
Sandberg 1982-1996 0.1883 -1.6536 0.0599
Sandberg 1996-2000 -0.0052 4.3000 0.3953
Sandberg 2000-2010 0.2144 1.6273 0.0855
4.1.2 Heat Wave Decadal Thresholds
The characteristics of heat waves for all six weather stations are summarized
within each of the following: Table 21, 22, 23, 24, 25, and 26. These tables describe
decadal trends for temperature thresholds T1 and T2, frequency, intensity, and the
average duration of heat waves for their respective station location over the 80 year
period (1931 to 2010). Specific heat wave trends and characteristics are described by the
following paragraphs.
The first weather station is Fairmont (Table 21) and it is important to note
that the decade spanning from 1931 to 1940 contain no daily temperature records and a
portion of the decade from 1941 to 1950 has missing temperature data. Also, the 20 year
span between 1991 and 2010 experience the highest T1 and T2 thresholds at 100.0°F and
89.9°F, respectively. The maximum and minimum number of heat wave events
(frequency) are 13 from 2001 to 2010 and 4 from 1941 to 1950, respectively. The
maximum and minimum intensity is 103.8°F from 1991 to 2000 and 100.2°F from 1941
to 1970, respectively. The final characteristic is an average duration of a heat wave event
78
and its maximum and minimum duration is 5.4 days from 1951 to 1960 and 2.1 days
from 1981 to 2000, respectively.
Table 21: Heat wave characteristics by decade from daily temperature data for the
Fairmont weather station from 1931 to 2010. T1 and T2 define the 97.5 percentile and 81
percentile of normal conditions, respectively. Frequency is the number of heat wave
events, intensity is the average maximum temperature of the heat wave, and average
duration is average span of each heat wave event.
Decade
Heat wave characteristics
T1 (°F) T2 (°F) Frequency Intensity (°F)
Average
duration (d)
Fairmont Weather Station
1931-1940 98.1 87.9 Incomplete Incomplete Incomplete
1941-1950 98.1 87.9 4 100.2 3.6
1951-1960 98.1 87.9 7 100.2 5.4
1961-1970 98.1 87.9 8 100.2 5.0
1971-1980 98.1 87.9 11 101.5 2.5
1981-1990 98.1 87.9 6 101.5 2.1
1991-2000 100.0 89.9 11 103.8 2.1
2001-2010 100.0 89.9 13 103.5 2.5
Average: 98.6 88.5 8.6 101.5 3.3
Another weather station with heat wave characteristics is Los Angeles (Table
22). Temperature threshold T1 and T2 reaches its maximum threshold conjointly from
1961 to 1980 with measurements of 92.1°F and 84.0°F, respectively. Additionally, the
maximum frequency is 14 heat events during two decades (1951 to 1960 and 1971 to
1980), and the minimum frequency is 3 heat events from 1941 to 1950. Maximum and
minimum intensity is 98.9°F during the 1971 to 1980 decade and 95.7°F from 1941 to
1950, respectively. The maximum average duration of a heat event is 4.3 days which
occurred from 1951 to 1960 and the minimum average duration of a heat event is 2.0
days and spans from 1941 to 1950.
79
Table 22: Heat wave characteristics by decade from daily temperature data for the Los
Angeles weather station from 1931 to 2010. T1 and T2 define the 97.5 percentile and 81
percentile of normal conditions, respectively. Frequency is the number of heat wave
events, intensity is the average maximum temperature of the heat wave, and average
duration is average span of each heat wave event.
Decade
Heat wave characteristics
T1 (°F) T2 (°F) Frequency Intensity (°F)
Average
duration (d)
Los Angeles Weather Station
1931-1940 91.9 82.0 4 95.9 4.0
1941-1950 91.9 82.0 3 95.7 2.0
1951-1960 91.9 82.0 14 95.5 4.3
1961-1970 92.1 84.0 9 98.2 3.2
1971-1980 92.1 84.0 14 98.9 2.5
1981-1990 92.1 84.0 18 98.6 3.0
1991-2000 92.1 82.9 10 97.7 3.0
2001-2010 92.1 82.9 8 96.9 3.0
Average: 93.2 82.9 10 97.2 3.1
The weather station at Palmdale (Table 23) demonstrates maximum T1
measurement of 104.0°F and T2 measurement of 96.1°F conjointly from 1991 to 2010.
Frequency reaches its maximum from 1951 to 1960 and 1991 to 2000 with 13 heat wave
events. The minimum number of heat events (frequency) occurs from 1941 to 1950 with
5 heat wave events. Maximum and minimum intensity measurements are 106.5°F from
1991 to 2010 and 104.9°F from 1961 to 1970, respectively. The final heat wave
characteristic is average duration with its maximum measurement is 6.7 days from 1961
to 1970 and its minimum measurement is 2.5 days from 1931 to 1940. Moreover, the
minimum measurement of 2.5 days from 1931 to 1940 must be considered with caution
because missing daily temperatures exist from January 1
st
, 1931 until March 31
st
, 1931
and October 1
st
, 1931 until August 15
th
, 1932. Furthermore, maximum measurements are
recorded in Palmdale for T1 (103.6°F), T2 (94.8°F), intensity (105.9°F), and average
duration (3.7 days) compared to all six weather stations. Therefore, these extremes show
that heat wave events are very intense and pronounced at the Palmdale location.
80
Table 23: Heat wave characteristics by decade from daily temperature data for the
Palmdale weather station from 1931 to 2010. T1 and T2 define the 97.5 percentile and
81 percentile of normal conditions, respectively. Frequency is the number of heat wave
events, intensity is the average maximum temperature of the heat wave, and average
duration is average span of each heat wave event.
Decade
Heat wave characteristics
T1 (°F) T2 (°F) Frequency Intensity (°F)
Average
duration (d)
Palmdale Weather Station
1931-1940 104.0 95.0 11 106.2 2.5
1941-1950 104.0 95.0 5 106.5 3.7
1951-1960 104.0 95.0 13 105.8 3.7
1961-1970 102.9 93.9 6 104.9 6.7
1971-1980 102.9 93.9 11 105.9 2.9
1981-1990 102.9 93.9 11 105.6 2.7
1991-2000 104.0 96.1 13 106.5 4.4
2001-2010 104.0 96.1 9 106.5 2.8
Average: 103.6 94.8 9.9 105.9 3.7
Pasadena (Table 24) temperature threshold T1 and T2 reach their maximum
from 1991 to 2010 with measurements of 98.9°F and 89.1°F, respectively. The
maximum and minimum number of heat events (frequency) are 20 from 1981 to 1990
and 7 from two temporally distant decades (1931 to 1940 and 1961 to 1970). Noteworthy
is the frequency trend from 1961 to 1990 where the number of heat events increases by
13 in a 30 year period. Also, the recorded number of heat events is the most in Pasadena
from 1931 to 2010 compared to all six weather stations which suggest that heat wave
events are experienced more here than any other weather station location. The maximum
intensity temperature is 102.7°F from 1991 to 2000 and the minimum intensity
temperature is 99.5°F from 1951 to 1960. Also, the maximum average duration of a heat
wave event is 4.1 days from 1951 to 1960 and the minimum average duration of a heat
wave event is 2.1 days from 1941 to 1950.
81
Table 24: Heat wave characteristics by decade from daily temperature data for the
Pasadena weather station from 1931 to 2010. T1 and T2 define the 97.5 percentile and
81 percentile of normal conditions, respectively. Frequency is the number of heat wave
events, intensity is the average maximum temperature of the heat wave, and average
duration is average span of each heat wave event.
Decade
Heat wave characteristics
T1 (°F) T2 (°F) Frequency Intensity (°F)
Average
duration (d)
Pasadena Weather Station
1931-1940 96.9 87.1 7 100.8 3.5
1941-1950 96.9 87.1 9 100.4 2.1
1951-1960 96.9 87.1 15 99.5 4.1
1961-1970 98.1 87.9 7 101.7 2.7
1971-1980 98.1 87.9 13 102.6 2.7
1981-1990 98.1 87.9 20 102.0 2.7
1991-2000 98.9 89.1 16 102.7 2.5
2001-2010 98.9 89.1 8 101.7 3.2
Average: 97.9 87.9 11.9 101.5 2.9
Sandberg (Table 25) most defining heat wave characteristics are described by
the following statistics. The decade from 1931 to 1940 contains no daily temperature
records and the decade from 1941 to 1950 contains approximately a year and half of daily
temperature data. Additionally, temperature threshold T1 and T2 reaches there maximum
from 1991 to 2010 with measurements of 93.9°F and 82.9°F, respectively. The minimum
frequency is 4 heat wave events, biased results must be considered due to the missing
daily temperature data, from 1941 to 1950; on the other hand, the maximum frequency is
10 heat wave events stretching 30 consecutive years (1971 to 2000). Also, the 80 year
frequency average is at the minimum in Sandberg with 7.7 heat wave events compared to
all six weather stations. This frequency statistic shows that heat wave events are less
frequent than any of the other six weather stations. The maximum and minimum
intensity measurements are 93.0°F from 1961 to 1970 and 96.4°F from 1991 to 2010,
respectively. Sandberg experiences the maximum average duration from 1961 to 1970
with a result of 5.8 days and a minimum average duration of 2.4 days from 1951 to 1960.
82
Table 25: Heat wave characteristics by decade from daily temperature data for the
Sandberg weather station from 1931 to 2010. T1 and T2 define the 97.5 percentile and
81 percentile of normal conditions, respectively. Frequency is the number of heat wave
events, intensity is the average maximum temperature of the heat wave, and average
duration is average span of each heat wave event.
Decade
Heat wave characteristics
T1 (°F) T2 (°F) Frequency Intensity (°F)
Average
duration (d)
Sandberg Weather Station
1931-1940 91.9 82.9 Incomplete Incomplete Incomplete
1941-1950 91.9 82.9 4 94.1 3.9
1951-1960 91.9 82.9 7 95.4 2.4
1961-1970 91.6 82.0 8 93.0 5.8
1971-1980 91.6 82.0 10 94.1 3.1
1981-1990 91.6 82.0 10 93.6 2.7
1991-2000 93.9 82.9 10 96.4 3.0
2001-2010 93.9 82.9 5 96.4 2.7
Average: 92.3 82.6 7.7 94.6 3.4
The final weather station discussed is the UCLA (Table 26) weather station
and this station has no daily temperature records from 1931 to 1940 and missing daily
temperature data for over a year and half of the collection period spanning from 1941 to
1950. Maximum temperature threshold T1 and T2 is recorded from 1991 to 2010 with
measurements of 89.9°F and 78.9°F. The maximum frequency is 12 heat wave events
and occurs from 1981 to 1990, and the minimum frequency is 1 heat wave event and this
occurrence is from 1941 to 1950. However, this minimum frequency result must be
considered biased due to the lack of daily temperature data spanning this decade. The
intensity of heat wave events reaches its maximum from 1981 to 1990 with a recorded
measurement of 94.8°F and a minimum recorded measurement of 90.9°F. Once again,
this minimum measurement needs to be considered with caution due to the lack of daily
temperature data for this time period. The average duration of heat wave events is at its
maximum from 1991 to 2010 at 3.8 days and is at its minimum with 2.4 days spanning
from 1991 to 2010. Overall, UCLA experiences the lowest T1 (89.2°F) and T2 (78.3°F)
83
measurements, the lowest intensity temperature (93.7°F), and a shared shortest average
duration of heat wave events (2.9 days) comparative to all six weather stations spanning
an 80 year time period.
Table 26: Heat wave characteristics by decade from daily temperature data for the UCLA
weather station from 1931 to 2010. T1 and T2 define the 97.5 percentile and 81
percentile of normal conditions, respectively. Frequency is the number of heat wave
events, intensity is the average maximum temperature of the heat wave, and average
duration is average span of each heat wave event.
Decade
Heat wave characteristics
T1 (°F) T2 (°F) Frequency Intensity (°F)
Average
duration (d)
UCLA Weather Station
1931-1940 89.1 78.1 Incomplete Incomplete Incomplete
1941-1950 89.1 78.1 1 90.9 3.0
1951-1960 89.1 78.1 9 94.5 3.3
1961-1970 89.1 78.1 11 93.7 2.6
1971-1980 89.1 78.1 10 94.3 3.8
1981-1990 89.1 78.1 12 94.8 2.6
1991-2000 89.9 78.9 7 93.7 2.4
2001-2010 89.9 78.9 7 94.5 2.4
Average: 89.2 78.3 8.1 93.7 2.9
An important note to mention is the mean results for T1, T2, frequency, intensity,
and average duration is not tested for statistical significance within this current study.
The difference of means test (t-test) would clarify the importance and significance of
these mean results. This test would measure the significance of differences between the
true mean and one or two sample means using the t-statistic result (University of Oregon
2014).
84
4.2 Monthly Temperature and Its Trend
The eight weather stations described in Table 4 are plotted to determine their
monthly temperature trend from approximately 1931 to 2010 (Figure 36, 37, 38, 39, 40,
41, 42, and 43). These eight figures illustrate that each station demonstrates a unique
trend over their respective time period and an increase in monthly surface temperature
over time. In addition, all of the stations exhibit a fluctuation in monthly surface
temperature due to seasonal variability. Seasonal variability can be explained as the
change in temperature from season to season; for example, winter months will experience
the coldest temperatures and summer months will experience the warmest temperatures
throughout the year. The following describes each weather station’s monthly
temperature trends in greater detail.
Fairmont (Figure 36) has a maximum monthly temperature of 85.6°F for the July
1
st
, 1931 measurement and a minimum monthly temperature of 31.3°F for the January 1
st
,
1937 measurement. These maximum and minimum monthly temperatures are evidence
that Fairmont experiences the largest range (difference between the maximum and
minimum monthly temperature) than any other weather station measuring monthly
surface temperature (54.3°F). Moreover, the period spanning from January 1
st
, 1981 to
December 1
st
, 1982 shows lower than normal monthly surface temperatures. These lower
monthly temperatures are ultimately caused by 13 months of missing monthly
temperature records during this two year time period.
85
Figure 36: Monthly surface temperature trend at the Fairmont weather station from 1931
to 2010. The bold blue line represents the monthly surface temperature and the bold
black line represents the trend line for monthly surface temperature. A single asterisk (*)
represents statistical significance at p-value < 0.05 and a double asterisk (**) represents
statistical significance a p-value < 0.01.
The LAX (Figure 37) weather station experienced normal seasonal monthly
temperature fluctuations, but an unordinary dip in December monthly temperatures
occurs in 1949 and 1950. The two lowest recorded monthly temperatures occur during
these two years with a temperature of 47.0°F in 1949 and 48.5°F in 1950. The warmest
monthly temperature on record at LAX is 76.5°F during the September 1
st
, 1984
measurement.
y = 0.0025x + 58.317
25
35
45
55
65
75
85
Monthly Surface Temperature ( °F)
Date
Fairmont
86
Figure 37: Monthly surface temperature trend at the Los Angeles International Airport
weather station from 1944 to 2010. The bold blue line represents the monthly
temperature and the bold black line represents the trend line for monthly surface
temperature. A single asterisk (*) represents statistical significance at p-value < 0.05 and
a double asterisk (**) represents statistical significance a p-value < 0.01.
Palmdale (Figure 38) experiences a normal seasonal trend of monthly
temperatures with the coldest temperature of 33.4°F recorded on the January 1
st
, 1937
measurement and the warmest temperature of 87.0°F recorded on the July 1
st
, 1931 and
July 1
st
, 1959 measurement. Additionally, biased results must be adhered to because 11
months of monthly temperature data is missing from April 1
st
, 1931 to December 1
st
,
1932.
y = 0.0027x + 60.284 **
25
35
45
55
65
75
85
Monthly Surface Temperature ( °F)
Date
LAX
87
Figure 38: Monthly surface temperature trend at the Palmdale weather station from 1931
to 2010. The bold blue line represents the monthly surface temperature and the bold
black line represents the trend line for monthly surface temperature. A single asterisk (*)
represents statistical significance at p-value < 0.05 and a double asterisk (**) represents
statistical significance a p-value < 0.01.
Pasadena (Figure 39) undergoes an increasing trend in monthly temperatures
during its respective time period. Monthly temperature records provide evidence of a
warming trend as seven of the monthly measurements have temperatures of 80°F or
greater since the August 1
st
, 1967 measurement with two of the station’s warmest
monthly temperatures of 81.6°F and 82.4°F takin place during the last 13 years;
respectively August 1
st
, 1998 and July 1
st
, 2006. A minimum monthly temperature is
recorded on the January 1
st
, 1937 measurement with a temperature of 43.4°F.
y = 0.0029x + 59.8
25
35
45
55
65
75
85
Monthly Surface Temperature ( °F)
Date
Palmdale
88
Figure 39: Monthly surface temperature trend at the Pasadena weather station from 1931
to 2010. The bold blue line represents the monthly surface temperature and the bold
black line represents the trend line for monthly surface temperature. A single asterisk (*)
represents statistical significance at p-value < 0.05 and a double asterisk (**) represents
statistical significance a p-value < 0.01.
Sandberg (Figure 40) experiences a wide range in minimum and maximum
monthly temperatures over the station’s time period. The maximum monthly temperature
recorded is 79.5°F on the August 1
st
, 1994 measurement and the minimum monthly
temperature is 27.1°F on the January 1
st
, 1949 measurement with this January
temperature the coldest on record compared to all eight weather stations. Important to
mention, the Sandberg monthly temperature data is lacking monthly temperature
measurements for 40 months spanning from August 1
st
, 1996 to October 1
st
, 2000. This
large gap in data explains the severe drop in monthly temperature illustrated in Figure 40.
y = 0.0049x + 60.452 **
25
35
45
55
65
75
85
Monthly Surface Temperature ( °F)
Date
Pasadena
89
Figure 40: Monthly surface temperature trend at the Sandberg weather station from 1948
to 2010. The bold blue line represents the monthly surface temperature and the bold
black line represents the trend line for monthly surface temperature. A single asterisk (*)
represents statistical significance at p-value < 0.05 and a double asterisk (**) represents
statistical significance a p-value < 0.01.
The results suggest that less extreme monthly temperature variations occur at
UCLA (Figure 41). The minimum monthly temperature at UCLA is 46.6°F and occurs
during the January 1
st
, 1937 measurement. Conversely, the maximum monthly
temperature experienced at this station is 77.2°F during the September 1
st
, 1984
measurement. Importantly, evidence suggests a shift in monthly temperatures since 1976
because UCLA experiences only three occurrences of monthly temperatures below 55°F;
however, prior to 1976 there are 19 occurrences of minimum monthly temperatures
below 55°F.
y = 0.0049x + 51.983 *
25
35
45
55
65
75
85
Monthly Surface Temperature ( °F)
Date
Sandberg
90
Figure 41: Monthly surface temperature trend at the UCLA weather station from 1933 to
2010. The bold blue line represents the monthly surface temperature and the bold black
line represents the trend line for monthly surface temperature. A single asterisk (*)
represents statistical significance at p-value < 0.05 and a double asterisk (**) represents
statistical significance a p-value < 0.01.
Another weather station that is experiencing less extreme monthly temperature
variations is USC (Figure 42). This evidence is illustrated in Figure 26 with its minimum
monthly temperature measured at 56.7°F and occurs on February 1
st
, 1950. USC’s
maximum temperature is recorded on September 1
st
, 1984 with a monthly temperature
measurement of 81.3°F. Additionally, this maximum monthly temperature measurement
is shared with UCLA and suggests that an occurrence of an accelerated temperature
increase is taking place at this time.
y = 0.0022x + 61.326 **
25
35
45
55
65
75
85
Monthly Surface Temperature ( °F)
Date
UCLA
91
Figure 42: Monthly surface temperature trend at the USC weather station from 1950 to
2010. The bold blue line represents the monthly surface temperature and the bold black
line represents the trend line for monthly surface temperature. A single asterisk (*)
represents statistical significance at p-value < 0.05 and a double asterisk (**) represents
statistical significance a p-value < 0.01.
Another station experiencing a subtle increase in their monthly temperature trend
is Woodland Hills (Figure 43) as compared to Pasadena’s more accelerated increase in
monthly temperature. Woodland Hills recorded a minimum monthly temperature of
45.6°F during the January 1
st
, 1950 measurement and a maximum monthly temperature
of 81.6°F during the August 1
st
, 1992 measurement. Also, larger seasonal temperature
variations are evident in Figure 43 with the largest range in seasonal temperatures
(30.8°F) occurring between the August 1
st
, 1992 measurement of 81.6°F and the
December 1
st
, 1992 measurement of 50.8°F.
y = 0.0027x + 63.308 *
25
35
45
55
65
75
85
Monthly Surface Temperature ( °F)
Date
USC
92
Figure 43: Monthly surface temperature trend at the Woodland Hills weather station from
1949 to 2010. The bold blue line represents the monthly surface temperature and the
bold black line represents the trend line for monthly surface temperature. A single
asterisk (*) represents statistical significance at p-value < 0.05 and a double asterisk (**)
represents statistical significance a p-value < 0.01.
4.2.1 Monthly Temperature Linear Regression Significance
The monthly temperature linear regression line is analyzed for its statistical
significance using the standard least squares regression analysis for all eight weather
stations and multiple statistics [i.e., coefficient of determination (R
2
), the slope
coefficient, and the p-value < 0.05 (95 percent level) or p-value < 0.01 (99 percent level)]
is categorized in Table 27. Also, the regression analysis sets the independent variable (y)
as date and the dependent variable (x) is set as monthly temperature.
Table 27 confirms that the linear regression line for the Fairmont, Palmdale, and
the Woodland Hills weather stations do not experience statistically significant monthly
temperature trends. The p-value and R
2
results are as follows: (1) Fairmont: p-value is
0.1148 and R
2
is 0.0027; (2) Palmdale: p-value is 0.0649 and R
2
is 0.0036; and (3)
Woodland Hills: p-value is 0.0596 and R
2
is 0.0048. On the other hand, Sandberg and
y = 0.0028x + 61.225
25
35
45
55
65
75
85
Monthly Surface Temperature ( °F)
Date
Woodland Hills
93
USC are statistically significant at the p-value < 0.05 level (95 percent level) with their p-
value and R
2
results as follows: (1) Sandberg: p-value is 0.0407 and R
2
is 0.0061 and (2)
USC: p-value is 0.0204 and R
2
is 0.0074. LAX, Pasadena, and UCLA are proven
statistically significant at the p-value < 0.01 level (99 percent level) with their p-value
and R
2
results as follows: (1) LAX: p-value is 0.0015 and R
2
is 0.0127; (2) Pasadena: p-
value is 0.0001 and R
2
is 0.0301; and (3) UCLA: p-value is 0.0005 and R
2
is 0.0129.
Overall, three weather stations (Fairmont, Palmdale, and Woodland Hills) are not
statistically significant per the standard least squares regression analysis. Two weather
stations are proven statistically significant at the 95 percent level and they include
Sandberg and USC. Lastly, three weather stations are statistically identified as
significant by the regression analysis and they include LAX, Pasadena, and UCLA.
Table 27: Statistical significance characteristics for monthly temperature at all eight
weather stations. A single asterisk (*) represents statistical significance at p-value < 0.05 and a
double asterisk (**) represents statistical significance a p-value < 0.01.
Weather
Station
Time Period R
2
Slope
Coefficient
p-value
Fairmont 1931-2010 0.0027 0.0025 0.1148
LAX 1944-2010 0.0127 0.0027 0.0015 **
Palmdale 1931-2010 0.0036 0.0029 0.0649
Pasadena 1931-2010 0.0301 0.0049 0.0001 **
Sandberg 1948-2010 0.0061 0.0049 0.0407 *
UCLA 1933-2010 0.0129 0.0022 0.0005 **
USC 1950-2010 0.0074 0.0027 0.0204 *
Woodland
Hills
1949-2010 0.0048 0.0028 0.0596
94
4.3 Yearly Temperature and Its Trend
4.3.1 Recorded Temperature at the Twenty-One Weather Stations
The 21 weather stations described in Table 3 are divided into a 20 year time
period (1931 to 1950) and a 60 year time period (1951 to 2010). These two time periods
are analyzed for two types of yearly temperature trends: 1) yearly surface temperature
and 2) yearly summer surface temperature; summer months are defined as the three
months of July, August, and September. Furthermore, the yearly temperatures are
obtained by averaging the monthly mean surface temperature dataset for each weather
station during their respective time period.
4.3.1.1 Recorded Temperature from 1931 to 1950
The trend analysis in Figure 44 illustrates the yearly surface temperature trends
for 20 weather stations in Los Angeles County. This figure reflects that the warmest
recorded yearly surface temperature occurring at the Palmdale station in 1931 with a
yearly temperature of 73.1°F. However, this extremely warm temperature must be
questioned due to Palmdale missing yearly temperature data during the coldest months of
the year (January thru March). These missing yearly temperature measurements would
thus lower the yearly temperature for 1931. Another obvious yearly temperature trend is
at the Table Mountain weather station. Table Mountain experiences the coldest yearly
temperature trend compared to all 20 weather stations. Hence, the coldest yearly
temperature recorded is 46.03°F in 1941 at Table Mountain.
Figure 44 illustrates a large fluctuation in yearly temperature between 1942 and
1945 at the OPIDS weather station. In 1942 a recorded yearly temperature of 55.3°F is
95
recorded and a yearly temperature of 57.68°F is recorded in 1945; a difference of 2.38°F
over a three year span which is the largest difference of all 20 weather stations. This
below average temperature is related to missing yearly temperature data from July to
October of 1937; therefore, this missing yearly temperature data explains the extreme
drop in yearly temperature over the average trend. Overall, the yearly surface
temperature trend shows low yearly temperature variability for most of the 20 weather
stations during the 20 year period, and this low variability suggests that neither a cooling
nor a warming trend exists for these weather stations.
96
Figure 44: Yearly surface temperature for the 20 weather stations in Los Angeles County,
California. The time period spans from 1931 to 1950 with the maximum and minimum
yearly surface temperature values recorded as compared to all 20 weather stations.
The next trend analyzed is yearly summer surface temperature for the same 20
weather stations from 1931 to 1950 (Figure 45). Figure 45 illustrates how these 20
stations yearly summer temperatures experience a higher variability of yearly
temperatures as compared to yearly surface temperature (Figure 44). Palmdale observes
the highest yearly summer temperature in 1937 with a recorded measurement of 80.0°F.
Moreover, Palmdale’s 20 year time period undergoes the warmest yearly summer
temperature trend compared to all other weather stations yearly summer temperature
measurements. Conversely, the coldest yearly summer temperature compared to all
45.00
50.00
55.00
60.00
65.00
70.00
75.00
80.00
85.00
1931 1936 1941 1946
Yearly Surface Temperature ( °F)
Year
Yearly Surface Temperature
(1931-1950)
Claremont Pomona Culver City Fairmont
Llano Long Beach Aquarium LA Terminal
LAX North Hollywood OPIDS
Palmdale Pasadena Pomona Fairplex
San Fernando San Pedro Sierra Madre Henszey
Table Mountain Torrance Airport UCLA
USC Valyermo Fire Department
Max Temperature: 73.1 ° F
Min Temperature: 46.0 ° F
97
weather stations is recorded at the Table Mountain station with a recorded temperature of
62.0°F in 1941. Another important discovery from yearly summer temperature is the
Sierra Madre Henszey weather station temperature increases by 4.47°F between 1941 and
1946, and this increase in yearly summer temperature suggests that a warming trend
exists during these five years. However, this warming trend must be questioned due to
the averaging process of monthly mean temperatures to calculate yearly summer
temperature.
Figure 45: Yearly summer surface temperature for the 20 weather stations in Los Angeles
County, California. The time period spans from 1931 to 1950 with the maximum and
minimum yearly summer surface temperature values recorded as compared to all 20
weather stations.
45.00
50.00
55.00
60.00
65.00
70.00
75.00
80.00
85.00
1931 1936 1941 1946
Yearly Summer Surface Temperature ( °F)
Year
Yearly Summer Surface Temperature
(1931-1950)
Claremont Pomona Culver City Fairmont
Llano Long Beach Aquarium LA Terminal
LAX North Hollywood OPIDS
Palmdale Pasadena Pomona Fairplex
San Fernando San Pedro Sierra Madre Henzsey
Table Mountain Torrance Airport UCLA
USC Valyermo Fire Department
Max Temperature: 80.0 ° F
Min Temperature: 62.0 ° F
98
4.3.1.2 Recorded Temperature from 1951 to 2010
Figure 46 illustrates six weather stations yearly surface temperature trend for a
total of 60 years. While all six stations are experiencing a warming trend, the most
pronounced warming trend occurs at the Pasadena weather station and LAX is
undergoing the most subtle warming trend of all six weather stations. The warmest
yearly surface temperature recorded is 68.89°F at the USC weather station, and a large
drop in yearly temperature occurs between 1997 and 1999 with a range of 4.18°F over the
three year period. Another important discovery is the coldest yearly temperature
recorded is 58.28°F in 1998 and 2009 at the Fairmont weather station. Also, large
portions of yearly temperature data is missing in 1956 and 1981 so these two year’s
temperature data is removed to avoid any skewness in the yearly surface temperature
trend.
99
Figure 46: Yearly surface temperature for the six weather stations in Los Angeles
County, California. The time period spans from 1951 to 2010 with the maximum and
minimum yearly surface temperature values recorded as compared to all six weather
stations.
Yearly summer surface temperature (Figure 47) illustrates some very different
trends for yearly summer surface temperature trends compared to yearly surface
temperature trends (Figure 46) from 1951 to 2010. The first difference in the temperature
trends suggests that a cooling trend exists for LAX and UCLA during their 60 year time
period. UCLA’s cooling trend is more accelerated compared to LAX’s more modest
cooling trend. Another difference between Figure 46 and Figure 47 is the maximum and
minimum yearly summer temperature measurements. Fairmont has the warmest recorded
yearly summer temperature in 2003 with a measurement of 82.13°F and the coldest
yearly summer temperature measurement is recorded at LAX in 2010 at 66.17°F. One
45.00
50.00
55.00
60.00
65.00
70.00
75.00
80.00
85.00
1951 1961 1971 1981 1991 2001
Yearly Surface Temperature ( °F)
Year
Yearly Surface Temperature
(1951-2010)
Fairmont Los Angeles International Pasadena UCLA USC Woodland Hills
Max Temperature: 68.9 ° F
Min Temperature: 58.3 ° F
100
similarity between the two yearly temperature trends is the accelerated warming trend at
the Pasadena weather station for the 60 year time period. Additionally, the most
accelerated cooling trend occurs in UCLA between 1984 and 1986 with a range of 6.94°F
during these three years. On the contrary, the most accelerated warming trend occurs
between 1982 and 1984 at Woodland Hills with a difference of 4.87°F during these three
years.
Figure 47: Yearly summer surface temperature for the six weather stations in Los
Angeles County, California. The time period spans from 1951 to 2010 with the
maximum and minimum yearly summer surface temperature values recorded as
compared to all six weather stations.
45.00
50.00
55.00
60.00
65.00
70.00
75.00
80.00
85.00
1951 1961 1971 1981 1991 2001
Yearly Summer Surface Temperature ( °F)
Year
Yearly Summer Surface Temperature
(1951-2010)
Fairmont Los Angeles International Pasadena UCLA USC Woodland Hills
Max Temperature: 82.1 ° F
Min Temperature: 66.2 ° F
101
CHAPTER 5: DISCUSSION & CONCLUSION
5.1 Extreme Temperature Threshold Observations
5.1.1 Weather Stations Experiencing an Accelerated Warming Trend
Three extreme temperature threshold variables (frost days, misery days, and heat
wave events) are analyzed and the results show that various trends are occurring at the six
weather stations. The evidence suggests an accelerated warming trend for the Palmdale
weather station and this warming trend is reflected by the pronounced increase in misery
days and decrease in frost days. Additionally, the misery and frost days linear regression
line is statistically significant at p-value < 0.01 or the 99 percent level. A possible
contributor to the resulting extreme temperature threshold trend and a suggested warming
trend in Palmdale is its geography.
For example, the Palmdale weather station is located in the high desert of
Southern California which experiences hot summer and mild winter temperatures. Also,
Palmdale is located on the leeward side of the San Gabriel Mountains and the orographic
effect can play a role on warmer temperature at these locales. The orographic effect
assists in warmer temperatures by the following: an accelerated descent of dry air
particles down the leeward side of a mountain leads to primarily cloudless skies; thus, an
increase in incoming solar radiation leads to a warmer surface temperature.
Furthermore, an investigation into the spatial discontinuity at Palmdale’s six
different daily temperature measuring locales reveals that high temperature variability
exists at this weather station. This high variability is greatly impacted by the station’s
surrounding physical features (i.e., type of landscape, adjacency to heat driven objects,
102
and the location of field measurements: near the surface or at the top of a building), the
distance between measuring locations (over 3,000 meters at maximum distance), and the
measuring locations are not in highly urbanized areas which reduces the impact of the
urban heat island effect.
Extreme temperature threshold results suggest another warming trend is occurring
at Fairmont during its respective time period. The warming trend, as compared to
Palmdale, experiences an accelerated increase in misery days and decrease in frost days.
A statistically significant linear regression line (p-value < 0.05 or the 95 percent level) for
misery days shows that this accelerated increase of misery days is occurring and
impacting Fairmont. Also, further evidence of a possible warming trend is documented
when heat wave records reveal that from 1991 to 2010 there are 24 heat wave events with
the most occurring since 1931 in the last decade (13 heat waves). Once again, geography
is a possible factor of the observed threshold trend at the Fairmont weather station
because it is located in the high desert of Southern California and its location on the
leeward side of the San Gabriel Mountains. Therefore, these warming trends over the last
80 years is an indication of a changing climate at the Fairmont and Palmdale weather
station
5.1.2 Weather Station Experiencing an Accelerated Cooling Trend
On the contrary, the extreme temperature threshold analysis identifies a cooling
trend occurring at the Sandberg weather station. This cooling trend is identified by an
accelerated increase in frost days and decrease in misery days. A statistically significant
linear regression line at p-value < 0.01 exists for misery days. The geographic location of
103
Sandberg is a possible contributor to this suggestive cooling trend and is described by the
following.
The temperature threshold influences at the Sandberg weather station is possibly
related to its high elevation at over 4,500 feet above sea level. Another suggestive factor
is Sandberg’s geographical location within an inland desolate, mountainous area. This
inland location is important on daily temperature variations because more intense
temperature fluctuations occur over land as compared to large bodies of water. These
large temperature fluctuations occur because energy or heat is absorbed rapidly at the
Earth’s surface and in turn large temperature changes occur over land. Another influence
from Sandberg’s location is the impact of urbanization is greatly reduced by negating the
urban island effect.
However, an extreme decreasing shift in misery days starting in 1975 requires
further research to determine the significance and cause of this decreasing extreme
temperature threshold shift. The study of the four spatially distributed daily temperature
measuring locations reveals high temperature variability at all four locations. Also, the
linear regression line during the extreme temperature threshold shift starting in 1975 is
not statistically significant which suggests that these threshold shifts are related more to
the measuring locations surrounding physical features and geography then to actual
extreme temperature threshold trends occurring at Sandberg.
104
5.1.3 Weather Stations Experiencing a Modest Warming Trend
Los Angeles, Pasadena, and UCLA extreme temperature thresholds are
experiencing a more modestly increasing warming trend, while Fairmont, Palmdale, and
Sandberg show stronger deviations of extreme threshold trends from historical
conditions. This indicative modest warming trend is suggested by the increase in misery
days and a more modest decrease in frost days. Also, the linear regression line for misery
days is statistically significant for Los Angeles and Pasadena at the p-value < 0.05. On
the other hand, Pasadena and UCLA experience a statistically significant linear
regression line for frost days at p-value < 0.01 and p-value < 0.05, respectively. Some
possible explanations for these three weather stations extreme temperature threshold
trends are discussed by the following.
This modest warming trend at these three weather stations is partially explained
by their close proximity to the Pacific Ocean. The close proximity to this large body of
water contributes to the moderation of temperatures at these weather stations. Another
reason for these moderate conditions is their low lying elevations with the average
elevations ranging at a minimum of 275 feet in Los Angeles and a maximum of 863 feet
in Pasadena (Table 2). Moreover, one might expect the urban heat island effect to play a
larger role on extreme temperature thresholds; thus, an accelerated warming trend at
these three weather stations would occur. This unexpected warming trend introduces the
possibility that other climate factors (internal or anthropogenic) are playing a larger role
in daily temperature observations and extreme temperature thresholds.
A deeper investigation into the spatial discontinuity at the six daily temperature
measuring locations for Los Angeles discovers that there is less threshold variability
105
compared to the six Palmdale and four Sandberg locations. The linear regression line for
misery days indicates statistical significance at the p-value < 0.01 level from 1940 to
1964 and no statistical significance for any of the six linear regression lines for frost days.
Once again, these results suggest that extreme temperature thresholds are highly
susceptible to the weather station’s physical surroundings and geography. Furthermore,
the result from the spatial discontinuity analysis in Los Angeles helps clarify the former
assumption that internal climate variability or anthropogenic forcing is playing a larger
role than the urban heat island effect.
5.2 Monthly Surface Temperature Observations
Monthly surface temperature shows an increasing trend in monthly surface
temperature for all eight stations analyzed and is an excellent data source to represent
seasonal temperature variability occurring at the eight weather stations. The largest
seasonal temperature fluxes occur at the Fairmont, Palmdale, and the Sandberg weather
stations and are attributed to their geographical location. Fairmont and Palmdale are
located in the high desert of Southern California with its characteristics defined as hot
and dry summer temperatures as well as relatively mild winter temperatures. Also,
Sandberg is located at a high elevation of 4,500 feet above sea level which produces
warm summer months and cold winter months. The standard least square regression
model affirms that linear regression line is not statistically significant for Fairmont and
Palmdale, but is statistically significant at the p-value < 0.05 level for Sandberg. The
large temperature fluxes at Fairmont, Palmdale, and Sandberg can be attributed to their
106
location in remote, mostly rural regions which greatly reduces the impact of the urban
heat island effect.
Pasadena is indicative of a warming trend (statistically significant: p-value < 0.01)
and high seasonal variability from the linear regression model and suggestive reasoning is
explained by the following three characteristics. Pasadena’s inland location limits the
effect of the moderate coastal climate and warmer temperatures that are experienced at
this weather station location. Another characteristic is the Mediterranean climate zone
that Pasadena is located within. The Köeppen climate classification of Cs, or a
Mediterranean climate zone, helps explain the extreme warming trend with very hot
summer days and mild winter days (George 2014). The last characteristic of Pasadena is
its urban locale because urban growth and urban landscape are playing a role in its
increase in monthly surface temperature.
LAX, UCLA, and USC are three stations that their linear regression line is
statistically significant at the p-value < 0.01 level and p-value < 0.05 level, and the linear
regression model indicates very modest seasonal variability for monthly temperature.
These stations are geographically located close to the Pacific Ocean which moderates the
seasonal variability and they are located within highly urbanized regions which increases
temperatures due to the urban landscape and influences monthly surface temperatures and
seasonal variability. Temperatures within urban landscapes will not experience the daily
temperature variability that is experienced in the rural surrounding thus moderating
monthly and seasonal temperatures. Overall, the modest warming trend introduces the
possibility that other climate factors (internal and anthropogenic) are playing a larger role
in dictating monthly temperature trends.
107
Woodland Hills is a prime example why monthly temperature is not an
appropriate temperature data source to analyze these temperature trends. The linear
regression model indicates that Woodland Hills is experiencing a very subtle increasing
warming trend over its respective 71 years, but its geographic location would suggest a
different monthly temperature trend. The geographic location of Woodland Hills is
within the Mediterranean climate zone and historical temperature records confirm that
very warm seasonal temperatures typically occur at this weather station. An example like
Woodland Hills proves that monthly surface temperature data is not a sufficient dataset to
identify a true representation of the occurring temperature trend, so the need to use daily
surface temperature data is required to represent the true historical temperature trends.
5.3 Yearly Surface Temperature Observations
Originally, the study’s goal was to collect as much yearly surface temperature
data and for as many stations in Los Angeles County spanning as many years as possible.
However, research and collection of historical temperature data clarified that extensive
yearly temperature data is limited and only one station in Los Angeles County had more
than 80 years of temperature data (Los Angeles Civic Center). In response to this limited
historical temperature data, the yearly surface temperature study is limited to 80 years.
These 80 years are subdivided into 20 (1931 to 1950) and 60 (1951 to 2010) year time
periods, and these two time periods are analyzed for yearly surface temperature and
yearly summer surface temperature. Analyzing these various time periods reveals that
yearly temperature trends show some explainable consistency such as warming trends
during specific time periods.
108
One explainable consistency is Table Mountain’s coldest yearly surface and
yearly summer surface temperature trend from 1931 to 1950. These cold temperatures are
explained by the elevation at Table Mountain with an approximate elevation of 7,500 feet
above sea level. This extremely high elevation keeps temperatures lower in the summer
months, as compared to temperatures in lower elevations, and frigid temperatures in the
winter months. Another consistency occurs during the yearly summer surface
temperature trend analysis from 1951 to 2010. Fairmont, Pasadena, and Woodland Hills
typically demonstrate the most extreme warming trends during these 60 years. These
stations locations within the high desert, further inland, and within a Mediterranean
climate zone help explain this yearly summer temperature trend. Also during this same
time period, LAX and UCLA experience a cooling trend over 60 years. This cooling
trend can be related to their proximity to Pacific Ocean as this large body absorbs and
loses energy or heat very slowly. This heat absorption process reflects the year round
moderate temperatures and less frequent extreme heat events. However, LAX and UCLA
are located within urbanized areas which suggests that the urban heat island effect would
produce a warming trend instead of a cooling trend at these two weather stations.
5.4 Overall Surface Temperature Observations
Daily surface temperature is a sufficient dataset for analyzing extreme
temperature thresholds and for determining the type of linear temperature trend occurring
at the six weather stations. On the other hand, the monthly surface temperature dataset
identifies seasonal variability, but does not provide sufficient temporal evidence of
surface temperature trends for the eight weather stations. Additionally, the yearly surface
109
temperature dataset does not provide adequate evidence of many explainable linear
temperature trends or seasonal variations for the 21 weather stations. A majority of the
yearly temperature trends show unexplainable inconsistencies because of the averaging of
monthly temperatures to obtain yearly surface temperature measurements. This
averaging loses the temperature extremes that are revealed within the daily surface
temperature data. Thence, the analyzation of daily surface temperatures is the key to
obtaining a true trend of surface temperature data. Most importantly, daily surface
temperature is a required dataset for measuring temperature trends over time, and an
analytical comparison of monthly and yearly surface temperatures is required to point out
that these two datasets are not adequate enough to analyze temperature temporally.
5.5 Spatial Characteristics and Distribution of the Weather Stations
The spatial distribution of the six daily surface temperature weather stations are
scattered across the north-northwest and south-central portions of Los Angeles County
with their characteristics following. Fairmont and Palmdale are located in northern
portion of the county and within the high desert region. Also, Fairmont is located in a
rural area and Palmdale is in a modestly urbanized area. Sandberg is located in the
northwest portion of the county and within a desolate, highly elevated region of the San
Gabriel Mountains. Pasadena, UCLA, and Los Angeles are located in the south-central
portion of the county and are within highly urbanized regions of the county. Also, UCLA
and Los Angeles are in close proximity of the Pacific Ocean.
The spatial distribution for the eight monthly surface temperature weather stations
are spread across the north-northwest and south-southcentral portions of the county.
110
Fairmont, Palmdale, Pasadena, Sandberg, USC (Los Angeles), and UCLA are described
formerly in the daily surface temperature weather stations. Woodland Hills is located in
the southcentral portion of the county and physically located within a highly urbanized
campus area. LAX is located in the southern portion of the county and is located within
an urbanized area bordering the Pacific Ocean with the area containing buildings,
airplane hangars, and tarmac.
Lastly, the 21 yearly surface temperature weather stations are dispersed
throughout the entire portion of Los Angeles County. Mostly the stations are located in
the south-southcentral-southeastern and north-northeastern region of the county. Three
weather stations (Fairmont, Llano, and Valyermo) are located in rural to desolate regions
of the high desert. On the other hand, Palmdale is located in modestly urban region of
the high desert. Table Mountain is located in the northeastern region at a high elevation
in the San Gabriel Mountain. Southcentral located weather stations include Long Beach,
Torrance Airport, LAX, and Culver City with all four of these stations located close to
Pacific Ocean and within modestly to highly urbanized areas. Two southeastern stations
include Claremont Pomona and Pomona Fairplex, and these stations are on the windward
side of the San Gabriel Mountains within highly urbanized areas. The remaining stations
within the southcentral portion of the county include: Woodland Hills; San Fernando;
North Hollywood; OPIDS; Sierra Madre Henszey; Pasadena; Los Angeles Terminal;
USC; and UCLA. All of these stations, except for OPIDS which is highly elevated
within the San Gabriel Mountains, are located at low lying elevations along the windward
side of the formerly mentioned mountain range. Also, all of these weather stations are
located within highly urbanized areas of Los Angeles County.
111
5.6 Current and Future Approaches
5.6.1 Current Study Advantages and Disadvantages
There are several advantages of the approaches used within the current
temperature analysis and they include the following. First, the extreme temperature
thresholds are prime indicators of warming or cooling trends over a long period of time at
any specific location. Also, the decadal trends of extreme temperature events show the
extreme temperature trends at certain time periods. Another positive approach is the use
of monthly temperatures to monitor the trends of seasonal variability. Lastly, the aid of
ArcGIS and Excel greatly reduces manual computations by performing tedious and
difficult algorithms and calculations for the location of weather stations as well as various
temperature calculations (i.e., yearly temperature, linear temperature trends, linear
regression line of temperature, etc.). There are definite advantages to the approaches
used within this study, but certain disadvantages are discovered and they are as follows.
The first disadvantage is that yearly temperature does not consistently show a
definitive temperature trend for a majority of the weather stations and the use of daily
temperature is required to discover a trend over time. Another disadvantage is the time
consuming and tedious process of manually dividing the daily temperature WRCC data
into ten year increments. Additionally, the manual process of dividing the monthly
temperature NCDC data for each station is extremely time consuming and very tedious
work. Another time consuming and tedious manual method is the averaging of monthly
temperature data into yearly temperature and yearly summer temperature. These
disadvantages are important forethoughts and must be considered for any future studies.
112
5.6.2 Future Study
The current study reveals that additional linear trend and regression analyses are
needed to provide a better understanding and greater insight into temperature trends and
possible contributors that affect temperature. For instance, further research would
involve a larger study area that includes the six Southern California counties: Imperial,
Los Angeles, Orange, San Bernardino, San Diego, and Riverside (SCAG 2009).
Broadening the study area introduces more weather stations and in turn yields more
historical temperature data. Obtaining large amounts of historical temperature data will
greatly increase the true representation of temperature fluxes and their temperature trends
across the southern part of California.
The other research route includes a large-scale regression analysis for two
metropolitan weather stations in Southern California and the selection of the cities plays a
very important role. This important role of the regression analysis will determine if
population plays the main role in temperature changes at the chosen locations or does
other chosen parameters (i.e., maximum temperature, minimum temperature, land cover,
etc.) the key factor to temperature changes at each city. The two cities are selected by
these criteria: 1) population growth increasing in the last ten years (i.e., Irvine,
California); and 2) population growth is slowing or declining in the last ten years (i.e.,
Bakersfield, California). Hence, a regression analysis can provide solutions in order to
correlate a relationship between temperature trends and possible influences such as
population growth.
113
REFERENCES
Anderegg, W., J. Prall, J. Harold, and S. Schneider. 2010. Expert credibility in climate
change. Proceedings of the National Academy of Sciences 107: 12107-12109.
Beaumont, L., A. Pitman, S. Perkins, N. Zimmerman, N. Yoccoz, and W. Thuiller. 2011.
Impacts of climate change on the world’s most exceptional ecoregions.
Proceedings of the National Academy of Sciences 108: 2306-2311.
Bohr, G. 2009. Trends in extreme daily surface temperatures in California, 1950-2005.
Association of Pacific Coast Geographers 71: 96-119.
Booth, E., J. Byrne, and D. Johnson. 2012. Climate changes in western North America,
1950-2005. International Journal of Climatology 32: 2283-2300.
Camilloni, I. and V. Barros. 1997. On the urban heat island effect dependence on
temperature trends. Climatic Change 37: 665-681.
Cayan, D., E. Maurer, M. Dettinger, M. Tyree, and K. Hayhoe. 2008. Climate change
scenarios for the California region. Climate Change 87: S21-S42.
Center for Climate and Energy Solutions. 2014. About the center for climate and energy
solutions. Accessed June 26, 2014. http://www.c2es.org/about.
Cordero, E., W. Kessomkiat, J. Abatzoglou, and S. Mauget. 2011. The identification of
distinct patterns in California temperature trends. Climatic Change 108: 357-382.
County of Los Angeles. 2014. LAC geography & statistics. Accessed June 3, 2014.
http://www.lacounty.gov/government/geography-statistics.
Doran. P. and M. Zimmerman. 2009. Examining the scientific consensus on climate
change. Eos Transactions American Geophysical Union 90: 22.
Esri 2014 ArcGIS Desktop 10.2. Esri Press: Redland, California.
George, M. 2014. Mediterranean climate. Accessed November 16, 2014.
http://californiarangeland.ucdavis.edu/Mediterranean_Climate/.
Goodridge, J. 1992. Urban bias influences on long-term California air temperature trends.
Atmospheric Environment 26B: 1-7.
Hall, A., F. Sun, D. Walton, S. Capps, X. Qu, H. Huang, N. Berg, A. Jousse, M.
Schwartz, M. Nakamura, and R. Cerezo-Mota. 2012. Mid-century warming in the
Los Angeles region: Part I of the “climate change in the Los Angeles region”
project. University of California-Los Angeles: Los Angeles, California.
114
Hansen, J., M. Sato, R. Ruedy, K. Lo, D. Lea, and M. Medina-Elizade. 2006. Global
temperature change. Proceedings of the National Academy of Sciences 103:
14288-14293.
Howard, L. 1833. Climate of London. Harvey and Darton: London, England.
Intergovernmental Panel on Climate Change. 2007. Contribution of working groups I, II
and III to the fourth assessment report of the Intergovernmental Panel on Climate
Change. In Pachuri, R. K. and A. Reisinger (eds), Synthesis report. IPCC,
Geneva, Switzerland, 104 pp.
Karl, T., J. Melillo, and T. Peterson. 2009. Global climate change impacts in the United
States. Cambridge University Press: Cambridge, United Kingdom.
Kaufmann, R., H. Kauppi, and J. Stock. 2006. Emissions, concentrations, and
temperature: A time series analysis. Climatic Change 77: 249-278.
Kaufmann, R., H. Kauppi, M. Mann, and J. Stock. 2011. Reconciling anthropogenic
climate change with observed temperature 1998-2008. Proceedings of the
National Academy of Sciences 108: 11790-11793.
Marietta College. 2014. The Mediterranean biome. Accessed June 4, 2014.
http://www.marietta.edu/~biol/biomes/shrub.htm.
Meehl, G., and C. Tebaldi. 2004. More intense, more frequent, and longer lasting heat
waves in the 21
st
century. Science 305: 994-997.
Michaelsen, J. 2009. Mojave desert region physical geography. Accessed
November 11, 2014.
http://www.geog.ucsb.edu/~joel/g148_f09/readings/mojave/mojave_desert.html.
Mishra, V., and D. Lettenmaier. 2011. Climatic trends in major U.S. urban areas, 1950-
2009. Geophysical Research Letters 38: 1-8.
National Aeronautics and Space Administration. 2014a. Climate change: How do we
know? Accessed November 2, 2014. http://climate.nasa.gov/evidence/.
National Aeronautics and Space Administration. 2014b. Global surface temperature-
yearly average. Accessed June 4, 2014.
http://climate.nasa.gov/climate_resource_center/23.
National League of Cities. 2013. The 30 most populous cities. Accessed June 3, 2014.
http://www.nlc.org/build-skills-and-networks/resources/cities-101/city-factoids.
Office of Environmental Health Hazard Assessment. 2013. Indicators of climate change
in California. By Tamara Kadir et al. Sacramento, California: OEHHA.
115
Oke, T. 1982. The energetic basis of the urban heat island. Quarterly Journal of the
Royal Meteorological Society 108: 1-24.
Oreskes, N. 2004. The scientific consensus on climate change. Science 306: 1686.
RssWeather. 2014. Climate for Los Angeles County, California. Accessed June 3, 2014.
http://www.rssweather.com/climate/California/Los%20Angeles%20County/.
Ruddell, D., D. Hoffman, O. Ahmad, and A. Brazel. 2013. Historical threshold
temperatures for Phoenix (urban) and Gila Bend (desert), central Arizona, USA.
Climate Research 55: 201-215.
Schlesinger, W. 2011. Climate change. Interpretation 65: 378-390.
Southern California Association of Governments. 2009. Climate change and the future
of Southern California. By Dan Cayan et al. Los Angeles, California: SCAG.
State of California. 2011. List of worldwide scientific organizations. Accessed June 4,
2014. http://opr.ca.gov/s_listoforganizations.php.
Taha, H. 1997. Urban climates and heat islands: albedo, evapotranspiration, and
anthropogenic heat. Energy and Buildings 25: 99-103.
Tamrazian, A., S. LaDochy, J. Willis, and W. Patzert. 2008. Heat waves in Southern
California: Are they becoming more frequent and longer lasting?
APCG Yearbook 70: 59-69.
United States Census Bureau. 1995. Population of counties by decennial census: 1900 to
1990. Accessed December 7, 2014.
http://www.census.gov/population/cencounts/ca190090.txt.
United States Census Bureau. 2014a. State & county quickFacts: Cook County,
Illinois. Accessed November 1, 2014.
http://quickfacts.census.gov/qfd/states/17/17031.html.
United States Census Bureau. 2014b. State & county quickFacts: Los Angeles County,
California. Accessed November 1, 2014.
http://quickfacts.census.gov/qfd/states/06/06037.html.
United States Department of the Interior: National Park Service. 2014. Mojave Desert.
Accessed November 7, 2014. http://www.nps.gov/jotr/naturescience/mojave.htm.
United States Environmental Protection Agency. 2014. Climate change indicators in the
United States, 2014. Third edition. EPA 430-R-14-004.
http://www.epa.gov/climatechange/indicators.
116
University of Oregon. 2014. Difference of means test (t-test). Accessed December 13,
2014. http://geog.uoregon.edu/geogr/topics/ttest.htm.
Western Regional Climate Center. 2014. Information on the California climate tracker.
Accessed November 4, 2014.
http://www.wrcc.dri.edu/monitor/cal-mon/frames_version.html.
World Meteorolgical Organization. 2014. WMO brief. Accessed June 26, 2014.
http://www.wmo.int/pages/about/index_en.html.
Abstract (if available)
Abstract
Climate change is a global occurrence and is studied at multiple scales within Los Angeles County, California. Determining the type of surface temperature trend across Los Angeles County is best observed using historical daily, monthly, and yearly temperature data. Each type of historical temperature data is analyzed for various temperature and extreme temperature threshold trends: (1) thresholds of frost days (minimum temperature ≤ 32°F), misery days (maximum temperature ≥ 90°F), and heat wave events are examined at six weather stations
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Spatial and temporal patterns of long-term temperature change in Southern California from 1935 to 2014
PDF
Distribution of Sonoran pronghorn (Antilocapra americana sonoriensis) on an active Air Force tactical range
PDF
Calculating solar photovoltaic potential on residential rooftops in Kailua Kona, Hawaii
PDF
Disparities in food access: an empirical analysis of neighborhoods in the Atlanta metropolitan statistical area
PDF
Risk analysis and assessment of non‐ductile concrete buildings in Los Angeles County using HAZUS‐MH
PDF
Generating bicyclist counts using volunteered and professional geographic information through a mobile application
PDF
Preparing for immigration reform: a spatial analysis of unauthorized immigrants
PDF
A comparison of urban land cover change: a study of Pasadena and Inglewood, California, 1992‐2011
PDF
Investigating bus route walkability: comparative case study in Orange County, California
PDF
Closed landfills to solar energy power plants: estimating the solar potential of closed landfills in California
PDF
Using Landsat and a Bayesian hard classifier to study forest change in the Salmon Creek Watershed area from 1972–2013
PDF
Smart growth and walkability affect on vehicle use and ownership
PDF
Increase in surface temperature and deep layer nitrate in the California Current: a spatiotemporal analysis of four-dimensional hydrographic data
PDF
Validating the HAZUS coastal surge model for Superstorm Sandy
PDF
Redefining urban food systems to identify optimal rooftop community garden locations: a site suitability analysis in Seattle, Washington
PDF
Monitoring parks with inexpensive UAVs: cost benefits analysis for monitoring and maintaining parks facilities
PDF
Finding your best-fit neighborhood: a Web GIS application for online residential property searches for Anchorage, Alaska
PDF
Historical observations of wildlife in Kenya: a Web GIS application
PDF
Comparing Landsat7 ETM+ and NAIP imagery for precision agriculture application in small scale farming: a case study in the south eastern part of Pittsylvania County, VA
PDF
Community gardens for social capital: a site suitability analysis in Akron, Ohio
Asset Metadata
Creator
Reed, Dustin Dwayne
(author)
Core Title
Historical temperature trends in Los Angeles County, California
School
College of Letters, Arts and Sciences
Degree
Master of Science
Degree Program
Geographic Information Science and Technology
Publication Date
01/27/2015
Defense Date
12/02/2015
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
California,climate change,extreme temperature threshold,geographic information science,historical surface temperature,Los Angeles County,OAI-PMH Harvest
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Lee, Su Jin (
committee chair
), Ruddell, Darren M. (
committee member
), Vos, Robert O. (
committee member
)
Creator Email
dustinre@usc.edu,gisman1980@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-525371
Unique identifier
UC11297565
Identifier
etd-ReedDustin-3140.pdf (filename),usctheses-c3-525371 (legacy record id)
Legacy Identifier
etd-ReedDustin-3140.pdf
Dmrecord
525371
Document Type
Thesis
Format
application/pdf (imt)
Rights
Reed, Dustin Dwayne
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
climate change
extreme temperature threshold
geographic information science
historical surface temperature