Seasonality and Climate Change: A Review of Observed Evidence in the United States


December 2021
SEASONALITY AND
CLIMATE CHANGE
A REVIEW OF OBSERVED EVIDENCE
IN THE UNITED STATES
U.S. Environmental Protection Agency

-------
The concept of seasonality refers to recurring events or processes that are correlated with seasons, such as rising temperatures
at the end of winter, the blooming of wildflowers in spring, the onset of allergies during ragweed season, and leaf-fall in autumn.
Three aspects of seasonality that are impacted by climate change are:
O TIMING: changes in when a
seasonal event occurs
©
during the year
Diminished snowpack and
early season melting in
mountain ranges...
J
DURATION: changes in
how long a seasonal event
lasts during the year
Earlier spring thaw and later
fall frosts alter leaf and
bloom dates and extend
the growing season...
..shifts water availability
especially during peak
crop growing season...
...which includes a longer
ragweed pollen season,
especially in northern states..
V
V
...which can lessen crop
yields or require more
groundwater use.
Cfi-
W
.and a longer period when
people may experience
seasonal allergies.
VARIABILITY: changes in
—'lAHju—intensity and frequency
of seasonal events
A longer, warmer,
and more extreme
wildfire season...

r.
V
...can influence changes in
the frequency, intensity,
and size of wildfires...
V
...which can lead to more
destruction and harmful
exposure to smoke and ash.
eft
EPA'S CLIMATE CHANGE
INDICATORS
In addition to support from the scientific literature, this report draws on data and findings from EPA's Climate
Change Indicators. EPA works in partnership with more than 50 data contributors from various government agencies,
academic institutions, and other organizations to compile a key set of indicators related to the causes and effects
of climate change. These indicators also provide important input to the National Climate Assessment and other
efforts to understand and track the science and impacts of climate change. Learn more about EPA's indicators at
https://www.epa.gov/climate-indicators.
Seasonality and Climate Change: A Review of Observed Evidence in the United States
2

-------
FRONT MATTER
The Earth's climate is changing. Multiple lines of evidence show changes in our weather, oceans,
ecosystems, and seasonal events. This technical report summarizes the current state of the science
on observed changes related to seasonality in the United States and discusses how climate change
affects the timing and nature of seasonal events. The report uses several key indicators sensitive to
and related to seasonality as a framework for understanding the implications of a changing climate
over time. The indicators are based on long-term observations and reveal the many dimensions of
seasonal events, including critical connections between physical changes and biological responses.
The report provides examples of how changes in seasonality affect ecological and human systems,
as well as our everyday lives
ACKNOWLEDGMENTS
Overall coordination, development, author contributions to this report were made by scientists of
the U.S. Environmental Protection Agency's Office of Atmospheric Programs and ICF. Contractor
support for content development and report design and production was provided by ICF under
contract number 68HERH19D0029.
PEER REVIEW
The primary indicators compiled and used in this report have been independently peer reviewed
as part of EPA's Climate Change Indicators effort. In addition, this Technical Report, including the
technical supporting documentation, was peer reviewed by three external experts in a process
independently coordinated by Abt Associates and an EPA peer-review coordinator. EPA gratefully
acknowledges the following peer reviewers for their useful comments and suggestions: Kathy
Jacobs, Holly R, Prendeville, and Scott Steinschneider. The information and views expressed in this
report do not necessarily represent those of the peer reviewers, who also bear no responsibility for
any remaining errors or omissions.
The Appendix to this report provides more information about the peer review.
RECOMMENDED CITATION
EPA. 2021. Seasonality and Climate Change: A Review of Observed Evidence in the United States.
U.S. Environmental Protection Agency, EPA 430-R-21-002. www.epa.gov/climate-indicators/
seasonalitv-and-cli mate-change
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
CONTENTS
TABLE OF FIGURES 	5
INTRODUCTION	6
WHY FOCUS ON CLIMATE CHANGE AND SEASONALITY? 	6
CHANGES IN SEASONALITY SERVE AS A BAROMETER FOR CLIMATE CHANGE 	7
OBSERVED SEASONALITY CHANGES IN THE UNITED STATES 	8
SEASONAL CHANGES HAVE ASSOCIATED DOWNSTREAM EFFECTS 	10
SEASONALITY AND CLIMATE CHANGE: A HEALTH FOCUS	11
SYNTHESIS OF OBSERVED CHANGES IN SEASONALITY 	12
THEME #1. SEASONAL CHANGES IN TEMPERATURE AND PRECIPITATION 	12
THEME #2. EFFECTS OF WARMER, SHORTER WINTERS 	14
Indicators of melting and thawing landscapes 	16
THEME #3. BIOLOGICAL RESPONSES AND PHENOLOGY IN A WARMING WORLD 	18
Timing Matters: Indicators of phenological mismatch	22
THEME #4. SEASONALITY AND EXTREME EVENTS 	24
SUMMARY AND RELATED RESEARCH 	28
SUPPLEMENTARY INDICATORS OF SEASONAL CHANGES 	29
SEASON LENGTH 	30
SEASON TIMING 	31
SEASONAL VARIABILITY 	34
APPENDIX 	38
FIGURES AND METADATA 	38
LIMITATIONS AND UNCERTAINTY 	38
INFORMATION QUALITY	39
PEER REVIEW	39
LITERATURE CITED	40
Seasonality and Climate Change: A Review of Observed Evidence in the United States	4

-------
TABLE OF FIGURES
Figure 1. Observed Evidence of Changes in Seasonality Related to Timing	 9
Figure 2. Seasonal Changes and Downstream Relationships	10
Figure 3. Change in Temperature by Season in the contiguous 48 States, 1896-2020 	 13
Figure 4. Trends in April Snowpack and Timing of Winter-Spring Runoff	 15
Figure 5. Timing of each Year's Arctic Sea Ice Melt Season 	 16
Figure 6. Change in the "Ice-Off" Date, or Date of Ice Thawing and Breakup, for 14 U.S. Lakes, 1905-2019 	 16
Figure 7. Ice Breakup Dates at Three Alaskan Rivers	 17
Figure 8. Freeze-Thaw Conditions: Frost-Free Season, 1979-2019 	 18
Figure 9. Length of the Growing Season in the Contiguous 48 States, 1895-2020 	 19
Figure 10. Change in the Length of the Growing Season for each State in the Contiguous 48 States, 1895-2020 ... 19
Figure 11. Change in the Length in the Ragweed Pollen Season, 1995-2015 	 20
Figure 12. Change in Latitude of the Average Winter Bird Center of Abundance, 1966-2013 	 21
Figure 13. Changes in Heat Wave Frequency and Season, 1961-2019 	 24
Figure 14. Wildfire Activity: Burned Acres and Seasonality in the United States, 1984-2017 	 25
Figure 15. Change in the Number of Days in the Atlantic Hurricane Season, 1970-2020 	 27
Seasonality and Climate Change: A Review of Observed Evidence in the United States
5

-------
INTRODUCTION
This technical report summarizes observed changes
related to seasonality in the United States, discusses
how climate change affects the timing and nature of
seasonal events, and describes some of the related
implications of those changes. To accomplish this, the
report uses a subset of indicators based on long-term
observations to explore the interconnectedness of
seasonal changes, including the cascading effect of
physical climatic changes and downstream biological,
ecological, and social responses.
The report aims to summarize the current state of the
science related to historical changes in seasonality
and provide tangible examples of the ways in which
climate change is altering the nature of seasonal
events—and how these changes affect ecological and
societal systems. Examining indicators of seasonal
processes and systems sensitive to seasonality
provides a framework for better understanding the
implications of a changing climate through time.
WHAT IS SEASONALITY?
The concept of seasonality refers to
recurring events or processes that are
correlated with seasons, such as rising
temperatures at the end of winter,
the blooming of wildflowers in spring,
the onset of allergies during ragweed
season, and leaf-fall in autumn.
This report considers three aspects of
seasonality: 1) shifts in the timing of
seasonal events (e.g., the timing of
animal migrations); 2) changes in the
duration of seasonal events (e.g., the
length of the wildfire season); and 3)
changes in the variability of events and
processes that occur during certain
times of the year (e.g., the number of
major hurricanes during the hurricane
season, the extent and severity of
wildfires, or the prevalence and spread
of Lyme disease).
WHY FOCUS ON CLIMATE CHANGE AND
SEASONALITY?
Although the timing, duration, and intensity of the seasons vary naturally from year to year, climate
change is driving longer-term changes in seasonality and fundamentally altering the ways in which
humans and natural systems experience and interact with seasonal events.1 At the core of these
changes are increases in temperature. These changes in physical climate lead to wide-ranging
impacts such as warmer winters; precipitation patterns shifting from snow to rain; species shifting
the timing or location of their seasonal activities, such as migration and reproduction; geographic
expansion and outbreaks of pests; and increases in the likelihood or duration of extreme events
such as heat waves, hurricanes, and wildfires.1 Some of these impacts may be beneficial, such as
longer growing seasons for crops or reductions in winter heating fuel costs. An understanding of
how seasonality is changing will inform efforts to monitor, communicate, and prepare for these
climate impacts.
Seasonality and Climate Change: A Review of Observed Evidence in the United States
6

-------
Key takeaways:
¦	Climate-related changes in seasonality are evident and happening now in most regions of
the United States. Changes in seasonality are well documented in indicators and the scientific
literature and can be found across the United States at different scales.
¦	Changes to seasonality have wide-ranging impacts—both positive and negative—across
physical, ecological, and societal systems. For example, increases in temperature can reduce
seasonal snowpack and lead to early snowmelt in mountain regions that provide water to
downstream agricultural areas,2 ultimately affecting crops and produce.3 Warmer and shorter
winters may reduce the length of the ski season, but could extend the season for other
recreational activities such as boating. Understanding how seasonal events are linked across
sectors can help us plan for and address risks associated with the effects of climate change at
seasonal and longer time scales.4
¦	Some of the changes underway can lead to harmful impacts on human health. Summer heat
waves, which are becoming more intense and frequent, can cause heat stroke, respiratory
problems, and other health conditions. Prolonged wildfire and pollen seasons can lead to
unhealthy air quality and pose risks for people with allergies and asthma.
¦	Climate indicators are important in framing and documenting changes in seasonality.
Long-term observational records, such as from weather stations or streamflow gauges, are
necessary to develop indicators that summarize changes in seasonality. Indicators help us
see how variations in the timing of seasonal events are tied to long-term changes in physical
drivers such as temperature and precipitation.5 Regular data collection and monitoring must
be in place in order to continually to track these trends.
WHAT IS AN INDICATOR?
An indicator is a convenient, useful analytical
expression of data that represents the state
or trend of certain environmental or societal
conditions over a given area and a specified
period of time. Indicators of seasonality
can illustrate such changes as increases or
decreases, rates of change, and the magnitude,
timing, duration, peak, severity, and frequency
of seasonal events.
United States.6108
Climate change drives shifts in the timing of seasonal events, such as the formation of lake ice in
winter, as well as the increased variability of seasonal events, such as the severity of heat waves
in summer. Climate change also alters the duration and frequency of seasonal events, potentially
affecting societal risk and prolonging exposure to more extreme events within a season.6 For
example, the wildfire season in the western United States now starts earlier, extends later into fall,
and includes more frequent and destructive wildfires than in the past.7 In turn, all of these changes
in seasonality drive impacts and consequences across environmental and societal systems.
CHANGES IN SEASONALITY
SERVE AS A BAROMETER
FOR CLIMATE CHANGE
Seasonality is dynamic, and characteristics and events
associated with the seasons can vary naturally from year
to year. For example, some years have hotter summers
and snowier winters, and some hurricane seasons have
stronger and more destructive storms. Human-induced
climate change has a strong influence over this variability
and is fundamentally changing seasonality in the
Seasonality and Climate Change: A Review of Observed Evidence in the United States
7

-------
Climate change is often difficult to contextualize and address because it manifests over long
timescales. However, changes in seasonality provide tangible evidence of climate change occurring
over relatively short timescales—including within human lifetimes. Because we understand seasonal
responses and relationships to climate, analyzing changes in seasonality can aid our understanding
of how the impacts of climate change may unfold over longer timescales.
Observational records of seasonality are increasingly used to monitor and analyze climate change
and its many consequences. Because climate can fluctuate naturally over the short term, scientists
require long records of observations to detect influences of climate change on seasonality. For
example, weather and year-to-year natural climate variations, such as El Nino and La Nina events,
can significantly affect seasonal conditions, at least over periods of a decade or less.8 Therefore,
it is important to collect and study long observational records to detect climate-related shifts in
seasonality that are separate from natural variability. The indicators described in this report have
been chosen because of their ability to demonstrate these long-term changes.
A number of recent studies attribute observed changes in seasonality to human-caused climate
change. For instance, studies show that recent record summer temperatures and heat waves in
the United States, such as those in 2014 and 2016, were likely to occur in the presence of human-
caused greenhouse warming and extremely unlikely to occur in its absence.910 Similarly, climate
change has reduced the probability of cold weather extremes in many parts of the world,10 and
the number of seasonal cold snaps in the United States has decreased significantly since the early
1900s.6 These findings underscore the impact that climate change has already had in reshaping
seasonal phenomena in the United States.
Moreover, studies have long documented the interface between climate change, seasonality, and
ecosystems.5 Across ecosystems worldwide, climate change has impacted the timing of seasonal
life history transitions—such as juvenile development and
reproduction—for a range of organisms, including plants,
birds, mammals, fish, and others.11-14 The majority of
observations and evidence for these changes are drawn
from terrestrial systems, but evidence from marine
systems is growing.15
This report summarizes current knowledge related to
the implications and relevance of observed changes
in seasonality. While observed changes in seasonality
are linked to human-caused climate change, this report
does not explore attribution of changes in seasonality to climate change, nor does it try to quantify
cause and effect beyond relationships that have already been published in peer-reviewed literature.
OBSERVED SEASONALITY CHANGES IN THE
UNITED STATES
This report summarizes a variety of observational evidence and indicators across geographies to
build a composite, data-rich view of changes in seasonality in the United States and their relation
to climate change. EPA tracks and compiles indicators using the best available monitoring data,
prioritizing datasets that represent regional to national geographies and long time periods.
EPA's Climate Change Indicators in the United States are available online: https://www.epa.gov/
climate-indicators.
Changes in seasonality provide tangible
evidence of climate change occurring
over relatively short timescales and
within human lifetimes. Climate change
impacts the timing, duration, frequency,
and severity of seasonal events in ways
that are important to human society.
Seasonality and Climate Change: A Review of Observed Evidence in the United States
8

-------
Figure 1 illustrates examples of indicators and key evidence of changes in the timing of seasonal
events that have been observed in the United States across various geographies and time scales.
These indicators provide multiple lines of evidence for shifts in the timing of seasonal conditions
and events.
Indicators reveal that warming temperatures have shortened frost seasons, led snowmelt to
occur earlier in the year, and contributed to a decline in snowpack. Similarly, wildfire and heat
wave seasons have increased in duration and severity, impacting ecosystems, human health, and
economies. Leaf and bloom dates are occurring earlier than before, and the growing season has
extended to cover a greater portion of the year. Subsequent sections of this report explore how
changes in one season cascade across and impact events in other seasons (e.g., winter conditions
affect harvests in the following fall).
Observed Evidence of Changes in Seasonality Related to Timing
SEASONAL CHANGE
KEY EVIDENCE
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

^	 LAKE ICE THAW & FREEZE 	^
Lake ice is thawing earlier by 24 days; Lake ice is freezing later by 10 days since 1973.a
¦4	 TIMING OF SPRING SNOWMELT
Spring snow is melting 8 days earlier on average since 1940.b
^	 ALASKAN RIVER ICE BREAKUP
River ice is breaking up 3 days earlier on average since 1924.c
^	 SPRING SNOWPACK
Spring snowpack is declining; snowpack peaks 9 days earlier on average since 1955.d
	 ARCTIC SEA ICE MELT SEASON 	^
Arctic sea ice melt season has increased by 38 days since 1979.c
	 FROST-FREE SEASON
^	 WILDFIRE SEASON 	~
Frost-free season is growing longer by 4 days per decade since 1979.ce
Wildfire season is occurring earlier, based on wildfire activity since the 1980s.d
4	 LEAF AND BLOOM DATES
¦4	 CHERRY BLOSSOMS PEAK BLOOM
Leaf and bloom dates are occurring earlier since 1981.e
Cherry blossom peak bloom date is 6 days earlier since 1921/
^	 HEAT WAVE SEASON 	^
Heat wave season is 47 days longer than in 1961.9
RAGWEED POLLEN SEASON 	~
Ragweed pollen season is from 6 days to 3 weeks longer since 1995.h
<4	 GROWING SEASON 	~
Growing season is 2 weeks longer overall since 1985 (last spring frost earlier; first fall frost later).e
a Northeast, Northern Midwest c Alaska 9 Contiguous U.S. 9 50 large cities across the U.S.
b West, Northeast	d West	f Washington, DC h Central U.S.
Figure 1. Summary of illustrative examples of changes in seasonality across the United States. Solid bars represent the
approximate time of year over which the indicators typically occur, and arrows denote earlier (^-) and later (—>) shifts in
the season as shown by indicator datasets. Note that some of these examples are limited to specific geographic regions
(e.g., cherry blossom peak bloom in Washington, DC). AH indicators and data were sourced from the EPA Climate Change
Indicators, available at: https://www.epa.gov/climate-indicators.
Seasonality and Climate Change: A Review of Observed Evidence in the United States
9

-------
SEASONAL CHANGES HAVE ASSOCIATED
DOWNSTREAM EFFECTS
Changes in seasonality are not isolated, but rather linked across time, space, and systems. Physical
changes in seasonality (e.g., changes in average winter temperatures or changes in the length
of the snow season) can affect a wide range of ecological and societal processes.16 For example,
warmer winters make it easier for agricultural pests to survive between growing seasons and
create adverse conditions for crops that require winter chill hours to bear fruit the following
season.17-19 Figure 2 provides an example of how seasonal processes are connected, from physical
climatological changes to ecological and societal responses. In this example, warmer winters can
reduce mountain snowpack and the volume of runoff-fed streams and reservoirs that are used for
crop irrigation. These impacts can converge and compound to affect the agricultural sector.20
Seasonal changes and downstream relationships
Decreasing snowpack:
Warming winter and spring
temperatures cause a decline
in seasonal snowpack across
the United States.
Decreasing runoff: Lower
snowpack reduces runoff
and water availability in the
Spring and Summer.
Worsening agricultural
conditions: Decreasing runoff
stresses regional agriculture,
particularly in areas that already
experience water scarcity.
Decreasing food availability:
These seasonal changes reduce
crop yields, which negatively
mpact communities and
businesses that rely on them.
imparl
busines
I
t!
Temperatures Increase
Snowpack Decreases
Warming temperatures:
Since the beginning of the
20th century, surface
temperatures across the
contiguous United States
have, on average, increased
by approximately 3°F in the
winter, and 2°F in the spring.
V j
Figure 2. Seasonal processes are interlinked and changes in seasonality can have cascading impacts across physical,
ecological, and societal systems. Note that this example focuses on decreasing snowpack leading to decreased water
supplies, which is a regional issue for the western United States.
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
SEASONALITY AND CLIMATE
CHANGE: A HEALTH FOCUS
Through its influence on the timing, duration, and variability of seasonal events, climate
change can affect people's potential exposure to new and existing hazards that pose risks to
public health. For example:
Extreme Heat: As global temperatures continue to rise, summers are becoming hotter and
the heat wave season is lengthening. Summertime temperatures in many parts of the United
States have already increased to thresholds that are challenging for groups at higher risk for
heat-related illness and mortality.1-90 Warmer winters are expected to reduce the number of
illnesses and premature deaths from exposure to cold, although several studies suggest that
this reduction will be smaller than the increase in illnesses and deaths from extreme heat.90
Wildfire: Wildfires in the western United States have become increasingly severe and cover
broader areas. In some areas of the West, the wildfire season is also becoming longer. These
changes can increase people's exposure to the health hazards of wildfires, including poor air
quality that can contribute to respiratory illness.190
Vector-borne diseases: Climate change—especially changes to temperature extremes and
precipitation patterns—can expand the geographic range, affect the abundance, and lengthen
the activity period of disease-transmitting ticks, mosquitoes, and other vectors that spread
illnesses such as Lyme disease and West Nile virus infection.*90
Pollen Season: Allergy seasons are becoming longer and more intense in some parts of the
country due to expanded pollen seasons and higher pollen counts for common allergen
species such as ragweed, grass, oak, and birch, leading to respiratory health impacts.1-90
Water-borne illnesses: Warmer fresh and marine waters can affect the range and growing
season of harmful algae and bacteria, viruses, and parasites, increasing the risk of exposure to
toxins and waterborne pathogens.1-90
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
SYNTHESIS OF OBSERVED
CHANGES IN SEASONALITY
The sections below present a synthesis of the observed changes in seasonality in the United States
as highlighted in four illustrative themes, or discussion topics. These topics explore changes in
seasonality as well as the ways in which these changes are interconnected across time and space
and between systems. This synthesis relies on many indicator datasets developed by EPA, along
with supporting scientific evidence and insights from peer-reviewed literature to provide a fuller
understanding of changes and associated impacts.
Seasonal temperatures and seasonal precipitation patterns in much of the United States have
changed since the early 20th century, driving many downstream ecological and societal impacts.
Observed changes in seasonal temperature include both increases in average temperatures and
more frequent and severe extremes, such as heat waves. Changing precipitation patterns include
long-term changes in average seasonal conditions as well as increased seasonal variability, which
drives more intense periods of drought and flooding.
Annual mean surface temperatures across the contiguous 48 states has been increasing by an
average of 0.16°F per decade since 1901, with a greater rate of increase since 1980.121 Warming has
occurred in each season (Figure 3), with winters warming the most and experiencing temperature
increases of nearly 3°F over this same period (1896-2019).22
Seasonal temperature in the West and upper Midwest states have warmed the most, while
states in the Southeast have warmed the least or cooled slightly (Figure 3). Examining long-term
temperature records, researchers found that the length of the four individual seasons is changing in
terms of start and end dates. In the Northern Hemisphere, summers grew longer between 1952 and
2011, whereas winter, autumn, and spring grew shorter108
Additionally, the coldest part of the temperature distribution is warming: the annual number of
days with extreme cold temperatures (colder than the 5th percentile) has declined since the 1950s
(see here), and upward trends in winter temperature minima have amplified the effects of warming
average winter temperatures.
Although experiencing a less frigid winter may seem good to some people, warming winters
have far-reaching consequences, including decreased snowpack,23,24 fewer chilling hours for
cold-dependent crops,1825 and a longer mosquito season.26 The section below on early season
warming and downstream impacts discusses these and other consequences in more detail.
THEME 1
SEASONAL CHANGES
IN TEMPERATURE AND
PRECIPITATION
Seasonality and Climate Change: A Review of Observed Evidence in the United States
12

-------
Change in Temperature by Season in the contiguous 48 States, 1896-2020
Spring
Figure 3. Map showing total seasonal temperature change, by state, across the contiguous United States from 1896 to 2020.
Right: Bar chart indicating average annual temperature change, by season, across the contiguous United States from 1896
to 2020. Seasons are defined as winter (December (from the previous year), January, February), spring (March, April, May),
summer (June, July, August), and fall (September, October, November). Source: EPA (2020
Winter	Spring	Summer	Fall
Season
Total temperature change (°F):
Other seasons have also warmed. Summer heat waves have increased in frequency, intensity, and
duration and, in the 2010s, increasingly occurred during spring and fall compared to the 1960s,
lengthening the heat wave season.27 Climate change also increases high-humidity events, which
combine with extreme heat to compound human health impacts.28 Extreme heat events can have
large, immediate impacts on animals, ecosystems, and human communities.29,30 Theme 4 below on
seasonality and extreme events discusses these issues in more detail
Similar to temperature, precipitation patterns in the United States have changed with respect to
both seasonal averages and extremes. Annual average precipitation increased at a rate of 0.17
inches per decade between 1901 and 2015, due mostly to large increases during the fall season.31
Changes in annual precipitation differ across regions, with increases in the Northeast, Midwest,
and Great Plains, and decreases in the West, Southwest, and Southeast.1
Changes in seasonal temperatures and precipitation patterns can alter hydrologic systems, seasonal
biology, and extreme events, as well as the downstream impacts associated with these changes.
The following sections focus on these important seasonal connections in more detail.
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
THEME 2
EFFECTS OF WARMER,
SHORTER WINTERS
Although temperatures in all seasons have been increasing in the United States, winter
temperatures have increased at the highest rate. A warmer winter can have several implications,
including some that manifest in other seasons, such as spring and summer water availability and
the length of the agricultural growing season. This section explores how physical changes such
as warming and shorter winters can influence a range of impacts, including to ecosystems and
human society. The box on page 16 describes how glaciers and the seasonality of lake and river
ice—including dates of ice freeze and breakup—are responding to climate change and variability.
The warming climate affects precipitation and other hydrologic processes, which has driven
pronounced changes to snowfall and snowpack in the United States. With respect to snowfall,
the proportion of winter precipitation that falls as rain rather than snow has been increasing.
Observations across the contiguous 48 states show that nearly 80 percent of weather stations
experienced a general decrease in the proportion of precipitation falling as snow from 1949
to 2020.32
Seasonal snowpack—the amount of snow that accumulates and compresses on the ground—has
also decreased over time (see Figure 4).33 Recent snowpack declines are almost unprecedented in
the last millennium, largely a result of springtime warming.34 The western United States, especially,
has experienced widespread temperature-related reductions in snowpack, with annual peak
snowpack depth decreasing by approximately 4 inches from 1982 to 2018. In addition, long-term
measurements indicate the annual timing of peak snowpack has shifted earlier at most locations.
This earlier seasonal trend is especially pronounced in southwestern states such as Colorado,
New Mexico, and Utah. Across all stations, peak snowpack has shifted earlier by an average of
nearly eight days since 1982.24 Furthermore, observations suggest the annual cycle of snowpack is
narrowing, corresponding to a shrinking of the length of the winter season.35
Snowpack provides two-thirds of the inflow to the major reservoirs in the western United States.24
As snowpack decreases and snowmelt occurs earlier in spring, western water availability becomes
increasingly stressed, which can contribute to drought conditions.3637 For example, the historic
California drought from 2011 to 2017 was exacerbated by record low snowpack in the Sierra Nevada
mountains caused by limited precipitation and high winter temperatures.3839 Loss of snowpack has
also led to decreasing water resources in other areas. Reduced snowpack decreases the albedo
effect (i.e., reflectivity: snow has a high albedo and so reflects a large proportion of incoming
solar radiation) and increases evapotranspiration, which have combined to decrease mean annual
discharge in the Colorado River.40 In parts of the country where streamflow is strongly influenced
by snowmelt, winter-spring runoff is happening at least five days earlier than in the mid-20th
century at most streamflow gauges (Figure 4). Trends are based on the winter-spring center
of volume, which is the date when half of the total streamflow has passed by each streamflow
gauge. The trends observed in these data are consistent with observations of changes in seasonal
snowmelt and streamflow in different regions of the country 41-44
Water deficits due to reductions in snowpack and snowmelt have critical consequences for
human populations, as snowmelt-derived water is used for hydropower, agriculture, and drinking
Seasonality and Climate Change: A Review of Observed Evidence in the United States
14

-------
water supply.24 California invested $2.7 billion in 2018 in new water projects, in part to deal with
increasingly persistent drought conditions and the need for greater water security.45 Additionally,
western U.S. wildfire activity has been shown to strongly correspond with warming and earlier
snowmelt.7 In Alaska, the tinning of snowmelt is one of the key factors that determines the onset
of the fire season in the boreal forest.46
Trends in April Snowpack	Timing of Winter-Spring Runoff
Percent change:
• • •• ••#9
More than 5 to 10 days 2 to S days 2 days earlier to
-40 -20 0 0 20 40	60 >60	...	* ,
to 80 to-40 to-2o to20 to4o toeo to»o	10 days earlier earlier	earlier 2 days later
Figure 4. Warming winters are driving declines in snowpack and earlier snowmelt and runoff. Trends in April snowpack in the
western United States from 1955 to 2020 (left); timing of winter-spring runoff in the western United States from 1940 to 2018
(right). Source: EPA (202®, n "
Winter outdoor recreation—particularly in the
Northwest, Northern Great Plains, and Northeast
regions of the United States—is impacted
by warming winters. As a greater portion of
precipitation falls as rain rather than snow,
snowpack decreases, spring thaws occur earlier,
and fall frosts occur later, activities such as skiing
are limited to a shorter period in the winter and a
smaller geographic range.1 This can have serious
economic consequences: in the Northeast alone,
the winter recreation industry supports nearly
45,000 jobs and brings in $2.6-$2.7 billion in
revenue annually.1 Projections indicate that activities
that rely on natural snow and ice cover may not
be economically viable by 2100 except in the
northernmost latitudes of the United States
unless global greenhouse gas emissions are
severely curtailed.1
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
INDICATORS OF MELTING AND
THAWING LANDSCAPES
Northern regions are warming faster than the rest of the world 47 The melting and thawing of
frozen, ice-covered landscapes, such as the near-global retreat of glaciers, widespread loss
of lake ice around the Northern Hemisphere, and disappearing Arctic sea ice, provide some
of the most iconic imagery for documenting the impact of climate change.47 Many of these
changes occur over long timescales, corresponding to warmer seasonal temperatures and
changes in the amount, timing, and type of seasonal precipitation.
May 1
May 31
Start of melt season
I
Aug 29
Sep 28

t
End of melt season

2000
Year
Arctic Sea Ice
Over the past several decades,
Arctic sea ice has declined
rapidly, with the largest decreases
occurring in the summer and
fall. Arctic sea ice decline is,
in large part, driven by rising
temperatures and a longer ice
melt season. Figure 5 shows this
lengthening trend in the Arctic
sea ice melt season, with the blue
lines indicating the start and end of
each year's season, and the shaded
band delineating season length.
Other factors that affect Arctic
sea ice extent include fluctuations
in oceanic and atmospheric circulation and natural annual and decadal variability. Declining
sea ice negatively impacts Arctic ecosystems, and Arctic wildlife are already threatened by
declining birth rates and restricted
access to food sources because
of reduced sea ice coverage and
thickness.33 These ecological
impacts, as well as the loss of
the ice itself, are also restricting
the traditional subsistence
hunting lifestyle of indigenous
Arctic populations such as the
Yup'ik, Inupiat, and Inuit.
Length
of melt
season
Figure 5. Timing of each year's Arctic sea ice melt season. The shaded
band spans from the date when ice begins to melt consistently until
the date when it begins to refreeze. Source: EPA (2020),1,0
Detroit Lake
^-0 days
Lake Osakis
-0 days
Lake Superior at Bayfield
-24 days
• Shell Lake
-6 days
Moosehead Lake
-5 days
Lake ice cover
Across 14 northern U.S. lakes,
freeze dates are occurring later
than they did historically, by
about 0.5 to 1.5 days per decade.
Additionally, lakes are, for the most
part, thawing earlier in the spring,
with ice breakup dates occurring
up to 24 days earlier than they did
Lake Mendota
^-7 days
Lake Monona
-8 days

Cobbosseecontee Lake Damariscotta Lake
-lOdays ^ *^-,odays
Mirror Lake ^
-6 days"**
SebagoLake
Lake George -16 days
-7 days %
Otsego Lake
-3 days
Geneva Lake
-5 days
Change in ice thaw date:
0 Earlier • No change
Figure 6. Change in the "ice-off" date, or date of ice thawing and
breakup, for 14 U.S. lakes during the period from 1905 to 2019. All
of the lakes have red circles with negative numbers, which represent
earlier thaw dates. Larger circles indicate larger changes. Source:
EPA (2020)."'
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
110 years ago (Figure 6).48 Later freeze dates and earlier thaw dates are a product of warming
temperatures, as well as other climate factors such as cloud cover, heavy precipitation,
and wind. This is consistent with observations of declines in seasonal ice cover duration
and concentration in the Great Lakes region and throughout the Northern Hemisphere.4749
Early ice-out or thaw in the spring can impair water quality in lakes, contributing to anoxic
conditions (low oxygen levels) or higher-than-average phosphorus and chlorophyll levels later
in the summer months. The loss of
lake ice has cultural significance
for Native Americans, including
limiting spiritual ceremonies, fish
harvest, travel, recreation, and other
activities.50 The cultural importance
of lake ice is evident in long-term
records of lake ice that have been
maintained for centuries.47'49 51
Ice breakup in three Alaskan
rivers Rising air temperatures
are in part contributing to earlier
spring ice breakup in the Tanana,
Kuskokwim, and Yukon rivers in
Alaska,52 with nine of the 10 earliest
ice breakup dates of the Tanana
occurring between 1990 and present day and the earliest spring ice breakup in recorded
history occurring in 2019.53 At all three locations, the earliest breakup dates on record have
occurred within the last two years. Earlier spring ice breakup has important implications for
the communities that live along the Tanana and Yukon, as these rivers serve as centers for
transportation, fishing, and hunting, and fulfill basic needs of surrounding wildlife communities.
Furthermore, early thawing as well as thawing of permafrost has the potential to damage
surrounding/overlying infrastructure and ecosystems through destructive ice movement,
extreme flooding, and soil instability.48
April 5
April 20
» Tanana River
• Yukon River
¦ Kuskokwim River
3 May 5
May 20
June 4
1895 1905 1915 1925 1935 1945 1955 1965 1975 1985 1995 2005 2015 2025
Year
Figure 7. Ice breakup dates at three Alaskan rivers: the Tanana,
Kuskokwim, and Yukon. Source: EPA (2020).'m
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
THEME 3
BIOLOGICAL RESPONSES
AND PHENOLOGY IN A
WARMING WORLD
Many plants and animals are sensitive to and respond to changes in seasonal events (i.e.,
environmental drivers and cues such as temperature thresholds or snowmelt), and thus their
responses make for good indicators. This section examines the ways in which plant and animal
species are responding to ongoing climatological changes, such as shifts in temperature, frost,
snowmelt, and wet and dry seasons. Observations reveal in rich detail the many ways in which
ecosystems and the agriculture sector are affected by shifts in these seasonal patterns.
One such set of observations pertains to phenology. Phenology is the study of plant and animal
life-cycle events in relation to environmental drivers such as weather and climate, and phenologica!
data make quite useful indicators.54 Ecological plant studies often report phenological events as
occurring on an annual date that is tied to local environmental cues such as temperature or the
length of winter (also referred to as vernalization). Recent observations have linked climate-induced
changes in phenology to human health impacts, including changes in the duration of the pollen
allergy season; cultural and outdoor recreational events, such as the National Cherry Blossom
Festival; wildfire activity; and agricultural yield.45 The text box at the end of this section further
explores the implications of phenological changes in response to climate.
This report includes evidence of temperature increases and warmer winters leading to earlier
snowmelt and longer frost-free seasons. For example, the duration of the freeze-thaw season has
increased steadily in recent years (Figure 8) This indicator uses satellite measurements to track the
number of days per year in the contiguous 48 states and Alaska in which the ground is unfrozen.
Changes in unfrozen days and the seasonal freeze-thaw cycle are important ecologically, as
a reduction in frozen land surfaces can lead to greater vegetation growth and productivity.55
Figure 8. Left: Change in annual frost-free season in North America. Right: Length of the frost-free season in the contiguous
48 states compared with the historical average (1979-2019). For each year, the bar represents the number of days shorter or
longer than average. Positive numbers represent years with more frost-free days than average (red bars). Negative numbers
represent years with fewer frost-free days than average (blue bars). Source: EPA (2020).m
Freeze-Thaw Conditions: Frost-Free Season, 1979-2019
Change in annual unfrozen days:
Seasonality and Climate Change: A Review of Observed Evidence in the United States 18

-------
In addition to an overall increase in the number of annual unfrozen days, the United States is
experiencing a longer growing season. Since 1980, the last spring frost has occurred an average
of three-and-a-half days earlier than the long-term average, and the first fall frost has occurred
about three days later (Figure 8).55 Correspondingly, the average length of the growing season
has increased by more than two weeks in the contiguous 48 states since 1895, with a more rapid
expansion since the 1970s (Figure 9).57
Length
E
o
c
o
'•4—'
CO
>
O
Spring Frost Timing —Fall Frost Timing
of the Growing Season in the Contiguous 48 States, 1895-2020
10
E
o
c
o
'¦4—1'
cc
>
CD
Q
Figure 9, Deviation in the timing of spring and fall frost dates from 1895-2020 relative to the long-term average
(1895-2020). Bottom: Deviation in growing season length from 1895-2020 relative to the long-term average (1895-2020).
Grey shading denotes periods in which the length of the growing season is increasing over multiple decades, largely in
correspondence to earlier cessation of spring frost and later onset of fall frost. Source: EPA (2020)."
Regionally, the largest increases in
the length of the growing season
have occurred in the Southwest, in
areas sensitive to warming and where
exposure to frost is historically low
(Figure 10). For example, both Arizona
and California experienced nearly
50 more growing days in 2020 than
in 1895.58
Extended growing seasons result in
both positive and negative impacts on
plant species and agriculture. Studies
show that longer growing seasons
extend forage production into late fall
and early spring, which benefits farmers
Change in length of growing season (days):
0 10 20
Figure 10. Change In the length of the growing season for each
state in the contiguous 48 states from 1895 to 2020. Source:
EPA (2020):-'
Growing Season Length
long-term
increase in
growing season
15
10
5
0
-5
-10
1895
1915
1935
1955
1975
1995
2015
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
by lessening their need to produce reserves to support livestock in winter.59 At the same time,
warming is expected to diminish forage quality, which requires farmers to produce more forage for
their livestock.59 In addition, warmer winters have increased survival rates for insect pests, which
further stress crops and increase the need and costs for pesticides during the growing season.
The agricultural sector relies heavily on predictable seasonal patterns in temperature, precipitation,
and corresponding animal and crop life cycles. In western regions fed by snowmelt, such as off the
Sierras, Cascades, and Rockies, earlier snowmelt or rainy seasons shifts runoff and water availability
away from the peak of the growing season when it is needed most. Surface water shortages
are particularly damaging to crop yields and increase the need for supplementary groundwater
pumping and irrigation.60 Water shortages are increasingly compounded by extreme heat events
and drought.6162
Warming-induced stress on vegetation and crops can carry a hefty price tag. Many tree and vine
crops require a certain number of "chill hours" during the winter to properly flower and fruit;
climate change-induced warming has begun and is expected to continue to reduce the number of
winter chill hours, which can negatively impact crops.63 For example, in the winter of 2016-2017,
warmer weather contributed to an 80-percent loss in Georgia's peach industry the following
summer, with roughly two-thirds of those losses attributed to a lack of chill hours necessary for
the fruit development.25
In addition to impacting agricultural yields,
increases in the number of growing and
flowering days have direct implications
for human health. For example, the
season for ragweed pollen, one of the
most common environmental allergens
in the United States, grew longer at 10
of 11 study locations in the Midwest from
1995 to 2015 (Figure 11). These trends are
strongly related to changes in the length
of the frost-free season and the timing of
the first fail frost.6465 The trend in a longer
ragweed season is more pronounced
within northern states, which is consistent
with warming trends showing that higher
latitudes are experiencing more rapid and
intense warming 66 Higher temperatures
throughout ragweed pollen season also
enable ragweed plants to produce larger
quantities of more allergenic pollen.6467
Observed increases in growing degree days
also correspond with a longer grass pollen season and an earlier start date for oak and birch pollen
seasons.68 Together, these factors contribute to longer allergy seasons and higher pollen counts.
The health effects of a longer and more intense allergy season have already been observed, such as
with increased sales of over-the-counter allergy medicine, more emergency department visits for
asthma attacks, and an increase in hay fever cases.69 A recent study of pollen observations found
that advances in the initiation of the pollen season and increases in spring pollen integrals strongly
support a phenological seasonal shift to earlier in the year.70
+24 days
+25 days
I
+21 days
+18 days
+15 days
^^+15 days
+15 days
I
+25 days
>
+11 days
+6 days
Change in length
of pollen season:
# Decrease # Increase	• -1 day
Figure 11. Change in the length in the ragweed pollen season
at 11 sites in the Midwest between 1995 and 2015. Source:
EPA (2020).64
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
Additionally, a warming climate can enhance the risk of vector-borne diseases, such as Lyme
disease and West Nile virus, by increasing the range of suitable vector habitat. The incidence of
Lyme disease in the United States has been increasing due to many factors. Increased temperatures
can create a longer seasonal period of deer tick activity and expanded habitat (see the section on
Supplementary Indicators of Seasonal Changes).71 In 2010, Lyme disease ticks were confirmed in the
upper Great Lakes region at Isle Royale and nearby Grand Portage National Monument (just below
the U.S.-Canadian border) for the first time.72
A welcome sign of spring occurs when certain flowering trees and plants blossom—a phenological
event. Changes in the timing of these events has significance to society and tourism. The National
Cherry Blossom Festival draws more than 1.5 million visitors annually to Washington, DC, to
enjoy blooms from the city's 3,000 cherry trees. Each year, the National Park Service determines
the dates of the festival by estimating when the trees will be in peak bloom based on local
temperatures during the winter and early spring. The average peak bloom date has been April 4
since recordkeeping began in 1921. However, the peak bloom date has occurred before April 4 in 14
of the past 20 years, and peak bloom dates shifted earlier by nearly six days between 1921 and 2019
(see Supplementary Indicators of Seasonal Changes).73
Animals also respond to seasonal changes. For example, changes in seasonality influence
migration patterns and behavior for many animals. Bird migration patterns are influenced by many
environmental cues, including seasonal temperatures, and long-term observations show that many
bird species have change their wintering ranges. The average winter center of abundance of 305
North American bird species shifted northward by at least 40 miles between 1966 and 2013 (Figure
12),74 with nearly 50 bird species shifting northward by more than 200 miles. Similarly, observations
show that the average center of biomass for 140 marine fish and invertebrate species shifted an
average of 20 miles northward and 21 feet deeper from 1982 to 2018.75 Other species shifts are
explored in the text box below on phenological mismatch.
Change in Latitude of the Average Winter Bird Center of Abundance, 1966-2013
80
-20
1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Year
Figure 12. Change in latitude of the average winter center of abundance among 305 North American bird species,
1966-2013. The shaded band shows the upper and lower confidence intervals or range of values around the average,
based on the number of measurements collected and the precision of the methods used. Source: EPA (2020).74
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
Some species are benefiting from climate change, although this can come at a cost to others. One
such example is the southern pine beetle. These beetles are freeze-intolerant, and their range has
historically been restricted to areas south of New Jersey. However, warmer winters have allowed
southern pine beetles to expand their range into New York, Connecticut, Massachusetts, and
Rhode Island, with recent northern outbreaks partially attributed to a winter warming trend.7677
Models estimate that by midcentury, southern pine beetles could expand their range throughout
the northeastern United States and even into Canada.78 The expansion of this beetle into new
areas is an ecological and economic concern, as they can cause extensive mortality in pine forests,
disrupting ecosystems services, shifting forest structure, and threatening native biodiversity.7678
Similarly, mountain pine beetles pose a threat to pine forests in the western United States. Since
2000, these beetles have impacted roughly 10.3 million hectares in this region and are ranked as
the most damaging forest insect on the National Insect and Disease Forest Risk Assessment.79
When mountain pine beetle infestations lead to tree mortality, they also increase the amount of
fuel available for wildfires. Moreover, climate change is expected to lead to higher temperatures
and more frequent and severe droughts—conditions that increase the frequency and severity of
both beetle outbreaks and wildfires.80 Ultimately, mountain pine beetle outbreaks and wildfires
result in compounding impacts, creating conditions in the western United States conducive to
both increased forest mortality and more frequent wildfires, but also downstream impacts such
as reduced air quality, river pollution, and erosion.
TIMING MATTERS: INDICATORS
OF PHENOLOGICAL MISMATCH
The science of phenology studies periodic events in the life cycles of plants and animals
that are influenced by variations in climate. Due to climate change, seasonal interactions
among species that have historically been synchronized—such as the timing of caterpillar
eggs hatching to match budbreak of their host trees for food—may become un-synchronized
(i.e., occurring at different times from each other) as species respond differently to changes
in seasonality.81 This phenomenon is referred to as phenological mismatch (or trophic
asynchrony) and can disrupt population dynamics, ecosystems, and ecosystem services.82
Phenological mismatch has been observed with Pacific black brant (a type of goose) and
sedge (plants that are a food source for the geese) in Alaska. Due to warmer winters and
earlier springs, these migratory geese have been arriving in Alaska earlier each year. Their
earlier arrival means they feed on sedge earlier in the plant's life cycle, leading to reduced
biomass, reproduction, and genetic diversity in the plant population. In turn, these changes
have implications for a shift in the sedge system from being a summer-season carbon sink to
a carbon source.83
Because it can be difficult to quantify and attribute phenological mismatch in observational
studies, experimental approaches have also been undertaken to explore whether certain
biotic interactions would experience mismatch and what the consequences might be. Several
studies have shown that phenological mismatch between insect hatches and budbreak or
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
4
leaf availability of their host plant can be caused by higher temperatures, and that this may
constrain the insect populations.84
Generally either one or both species in the mismatched interaction will experience negative
consequences based on the nature of their relationship (e.g., antagonistic predator-prey
interactions may see an increase in abundance of the prey if they become mismatched
from their predator, whereas in a mutualistic interaction—such as with pollinators and
flowers—both species may suffer from decreased nutrition and reproductive success,
respectively).85 Although it can be difficult to predict how interspecies interactions and
ecosystem communities will shift under climate change, we can look to current examples
as illustrations of the types of changes that could occur more frequently in the future.
Phenological mismatches can also occur at the societal level, posing particular challenges
for Tribal and Alaska Native communities. For example, shifts in the timing of plant flowering
or fruit production, or in the timing of bird, mammal, or fish migration, can reduce the
availability of culturally significant foods and medicinal plants, make them available outside
of legal harvest windows, or create mismatches with traditional time-specific ceremonies or
other events. In Alaska, thinning sea ice in springtime creates dangerous conditions for seal
and walrus hunters.113
Seasonality and Climate Change: A Review of Observed Evidence in the United States

-------
THEME 4
SEASONALITY AND
EXTREME EVENTS
&
In addition to its impact on long-term average changes in seasonality, climate change can affect
the frequency and/or severity of shorter-term acute or extreme events with disproportionately
large-scale impacts. Extreme climatological events that are closely linked to seasons include heat
waves, wildfires, and tropical cyclone activity. In many cases, climate change is increasing the
frequency and severity of these seasonal extremes, as well as lengthening the period of the year
during which they occur.
Summer heat waves are becoming more intense and frequent, and the duration of the period of the
year experiencing heat waves has increased as a result of climate change.86 87 In this report, heat
waves are defined as a period of two or more consecutive days where the daily minimum apparent
temperature (actual temperature adjusted for humidity) exceeds the 85th percentile of historical
July and August temperatures (1981-2010) for that location. In the last half century, the United
States has seen marked increases in extreme high temperatures, with heat waves occurring almost
three times more frequently in recent years compared with the long-term average.27An analysis
of urban temperature shows 46 of 50 major cities in the United States experienced a statistically
significant increase in heat wave frequency between the 1960s and 2010s (Figure 13). The average
heat wave season (the number of days between the first heat wave of the year and the last)
increased by 47 days across the 50 cities over the same time period.27
Changes in Heat Wave Frequency and Season, 1961-2019
Heat Wave Frequency Heatwave Frequency
o «*
w ni
w
a
2 5
1960s 1970s 1980s 1990s 2000s 2010s
Heat Wave Season
(V flj
2000s 2010s
* •
Change in number of heat waves per year:
Not S2 2 to 4 4 to 6 6 to 8 >8
Heat Wave Season
• > >
• «•
.. .•
••
*> >
• • •
Change in season length (days):
Figure 13. Changes in the number of heat waves per year (frequency) and the number of days between the first and last
heat wave of the year (season length). A heat wave is defined as a period of two or more consecutive days where the daily
minimum apparent temperature exceeds the 85th percentile of historical July and August temperatures (1981-2010) for
that location. These data were analyzed from 1961 to 2019 for 50 large U. S. metropolitan areas. The graphs show averages
across all 50 metropolitan areas by decade. The size/color of each circle in the maps indicates the rate of change per
decade, Hatching represents cities where the trend is not statistically significant. Source: EPA (.2020X27
Seasonality and Climate Change: A Review of Observed Evidence in the United States
24

-------
Heat waves, especially when coupled with periods of high humidity, have consequences for
animals, ecosystems, and human communities.3031 Extreme heat events can induce a range of
dangerous—and sometimes fatal—human health conditions, including heat stroke, and have been
linked to respiratory problems as a result of harmful air pollutant build-up.88 In the United States,
extreme heat events cause more deaths annually than any other form of extreme weather.89
Vulnerable populations, including elderly people and people living in poverty, often experience
particularly negative impacts from extreme heat.90
Warmer conditions also contribute to more intense and prolonged wildfire seasons in the United
States. Changes in wildfire seasons are connected to rising temperatures through a range of
processes in conjunction with other human-induced factors.9192 For example, human activities
and land management practices also affect wildfire activity, and preferred practices in wildfire
management have evolved over time, from older policies that favored complete wildfire prevention
to more recent policies of wildfire suppression and controlled burns. Resources available to fight
and manage wildfires can also influence the amount of area burned over time. Warmer springs
and summers increase atmospheric aridity and generate more dry fuel and flammable forests that
feed large wildfires. In California, warming increases the likelihood that dry fuels persist into fall
when strong wind events, such as the Santa Ana winds, fan and spread wildfires over large areas.93
Compounding matters, heat waves can amplify these processes to precondition landscapes for
large wildfires. Declines in summer precipitation and wetting rain days (widespread rain that over
an extended period of time significantly reduces fire danger94) are also a key driver in the increased
extent of wildfires in many areas of the western United States.95
Since the late 20th century, the length of the wildfire season—the period between the first and
last large wildfire of the year—has increased in step with temperature increases. On average, fire
seasons from 2003-2012 were almost three months longer than those from 1973-1982.7 Wildfire
seasons have also become more severe, with more frequent and larger fires burning larger areas.
Notably, the five years with the greatest wildfire acreage burned since 1960 all occurred in the last
15 years, with burn extents during the 2015, 2017, and 2020 seasons exceeding 10 million acres
nationally.96 The increase in burn area has been most significant in western states (Figure 14),
where the area burned has more than doubled over the past 40 years in many states. While most
increases have occurred in the summer months during the height of the wildfire season, additional
increases have occurred in the spring and fall, signaling an extension of the wildfire season through
time. Meteorological observations used to estimate the length of fire weather season show
increases over a similar period.97
Wildfire Activity: Burned Acres and Seasonality in the United States, 1984-2017
Change in annual burned acres per square mile of land area:
m

1.8

(0
01
1.6

u
1.2
u

-------
Increasingly prolonged and more severe wildfire seasons have had devastating impacts on human
communities and ecological systems across the western United States. Tangible impacts, such as
the destruction of residences and city infrastructure, are often represented in dollar loss statistics
that indicate the "price" of wildfire burn in a given year. The total U.S. wildfire dollar loss nearly
doubled between 2010 and 2018, rising from $13.4 billion to $25.6 billion.98 The National Oceanic
and Atmospheric Administration tracks the number and cost of billion-dollar weather and climate
disasters (individual events where the cost of damages meets or exceeds $1 billion).1 Between 1980
and 2020, there were 18 billion-dollar wildfire seasons in the United States; 15 of those had occurred
since 2000." Eight of these billion-dollar wildfire seasons occurred between 2010 and 2020, with
the estimated total costs for these eight seasons totaling over $58 billion." Some of these damages
were influenced by increased development and land use changes during this time period.
The consequences of worsening wildfire seasons far exceed those represented by a dollar statistic.
For example, smoke from wildfires has significant air quality impacts: in 2020 alone, more than
50 million—or one in seven—Americans experienced "unhealthy" air from smoke during wildfire
season.100 In addition, wildfire smoke can significantly decrease photosynthetic activity, negatively
impacting plant communities far beyond immediate burn areas.101 Severe wildfires also have
detrimental impacts on local water quality. Post-burn soils are hydrophobic, meaning that they
repel water and increase surface runoff during storms. These factors increase the risk of mudslides
as well as the likelihood of debris and elevated levels of nitrogen and phosphorus entering water
systems and causing algal blooms.102
The degree to which climate change has impacted hurricanes and tropical storms, collectively
referred to as tropical cyclones, is a topic of intense interest given that their extreme rains, high
winds, and storm surge can cause severe property damage, soil erosion, flooding, and loss of life.
Tropical cyclones most commonly occur during the "hurricane season" running from June through
November and draw their energy from warm tropical oceans. Changes in sea surface temperatures
can alter the intensity of wind and rain associated with tropical cyclones.
Observations show an upward, non-significant trend in the frequency of hurricanes occurring in
the North Atlantic since the 1970s,103 which has been driven, in part, by reductions in aerosols from
human activity.104 The intensity and duration of tropical cyclones also increased during that period
as measured using the Accumulated Cyclone Energy (ACE) Index, which considers maximum wind
speeds throughout the lifetime of each observed storm.103 A recent study found that observed
increases in the intensification of hurricanes between 1982 and 2009 are highly unusual compared
to modeled internal climate variability and are consistent with a positive contribution from human-
caused climate change.105 However, it is premature to conclude with high confidence that climate
change has imparted a broadscale detectable impact on hurricane activity, because the quality of
the observed record is limited prior to 1970 and natural variability plays such a large role.
Researchers have also investigated whether the Atlantic hurricane season is lengthening. One
study found that the dates between the earliest and latest tropical cyclone formations in the
Atlantic were widening by about one day per year, potentially as a result of increasing sea surface
temperatures.106 Figure 15 shows the number of days between the first and last tropical cyclone,
hurricane, and major hurricane within the Atlantic basin for each calendar year since 1979. Similar
to other studies, this analysis shows that the length of the season has grown by about 1-2 days per
year over the recent period. However, the attribution of these trends to any single driving factor
is difficult and other studies have found conflicting results meaning that the uncertainty in these
trends is high.107
Seasonality and Climate Change: A Review of Observed Evidence in the United States
-------
Change in the Number of Days in the Atlantic Hurricane Season, 1970-2020
350 rr
Named storms
Hurricanes
Major hurricanes
Year
Figure IS. Number of days between first and last occurrence of named storms, hurricanes, and major hurricanes in the
Atlantic and linear trends for the period 1970-2020, Data source: National Hurricane Center HURDAT2 archive.
Seasonality and Climate Change: A Review of Observed Evidence in the United States
-------
SUMMARY AND RELATED RESEARCH
The systems and relationships described above provide multiple lines of evidence showing that
many changes in seasonality are consistent with a warming world. Tracking these observed changes
through indicators is one way in which we can better communicate and understand the broader
implications of a changing climate. Together, the indicators and other observational evidence
related to seasonality discussed in this report clearly demonstrate that such changes are occurring
now across all regions of the United States, and have far-reaching associated consequences to
ecosystems, human health and well-being, and the economy.
This report leverages several readily available data sources; however, additional monitoring, data,
and indicators are needed to improve our ability to characterize these changes and address gaps
in knowledge. For example, only limited data are available to understand and develop indicators
on changes in seasonality and phenology in marine and coastal environments. Several biological
and life-cycle processes are seasonal and based on environmental cues (e.g., fish migration,
phytoplankton blooms, spawning). Similarly, more comprehensive and widespread sampling data
on pollen in the United States are needed to better characterize seasonal changes in pollen and
allergies for the purposes of protecting public health.
The climate research community is increasingly interested in research at seasonal-to-sub-seasonal
scales and climate-sensitive processes relevant to such areas as public health and national security.
Predictive models are often used to estimate the severity of extreme heat in the upcoming
summer, the spread of infectious diseases in winter, or how mosquito populations will respond to
precipitation and temperatures in spring. Indicators provide insights into the historical behavior of
these seasonal processes, which can be assimilated into predictive models or used to benchmark
the severity of forecasted outcomes. Changes in the timing and nature of seasonal processes will
require adjusting policies and management practices to adapt to new conditions. Thus, combining
historically based data and indicators with predictive modeling will improve our ability to
understand and plan for these changes.
Seasonality and Climate Change: A Review of Observed Evidence in the United States
-------
SUPPLEMENTARY INDICATORS OF
SEASONAL CHANGES
This section presents an expanded set of EPA's climate change indicators relevant to seasonality
to supplement the indicators and figures presented in the main report and to demonstrate the
breadth and depth of seasonality changes occurring in the United States and around the world.
The following pages show figures and key points for each supplementary indicator. The indicators
include a range of temporal and spatial scales, and it is best to interpret these changes individually.
The summary table below summarizes all of EPA's climate change indicators features in the main
report and this supplementary section. For reference, the indicators are grouped into the general
dimensions of seasonality discussed in the report, including seasonal timing, length, and variability.
For background information and additional figure details for all of these indicators, see the EPA
Climate Change Indicators webpage.
Table I: Summary of EPA's Climate Change Indicators Related to Seasonality.
Bolded indicators are those that appear in the supplementary section below.
SEASON LENGTH
SEASON TIMING
SEASONAL VARIABILITY
1. Great Lakes Ice Cover
1. Peak Snowpack Date
1. Heating and Cooling

Degree Days
2. Arctic Sea Ice Melt Season
2. High Winter-Spring

Flow Carried by Rivers
2. Snowfall
3. Hurricane Season
and Streams

3. Seasonal Temperatures
4. Freeze-Thaw Season
3. Alaskan River Ice Breakup
(Unfrozen Days)
4. Glacier Mass Balance

4. Lake Ice

5. Growing Season
(Freeze-thaw Dates)
5. Residential Energy Use

(summer and winter)
6. Growing Degree Days
5. Timing of Spring Snowmelt

6. Lyme Disease Prevalence
7. Heat Wave Season
6. Leaf and Bloom Dates

7. Snow-to-Precipitation Ratio
8. Ragweed Pollen Season
7. Cherry Blossom
9. Wildfire Activity: Season
Bloom Dates
8. Wildfire Activity: Extent
Seasonality and Climate Change: A Review of Observed Evidence in the United States
29
-------
SEASON LENGTH
Key Point: Since 1973, ice cover on Lakes
Huron, Michigan, Ontario, and Superior
has decreased at rates ranging from
approximately one-quarter of a day to
more than half a day per year.
Trend in days per year:
^ j I I I 1 I 1 1 No trend
-2.5 -2.0 -1.5 -1.0 -0.5 0 0.5 1.0 1.5 2.0
Average annual rate of change in the duration of ice cover in the Great Lakes from 1373-2019.
Duration is measured as the number of days in which each pixel has an ice cover concentration of at
least 10 percent.
• • * -
w • . • v ."-War
% • • •
• •••

• •

• •
Key Point: Between 1994 and
2018, the number of growing
degree days increased in all
regions of the contiguous 48
states. The largest increases
occurred in the Southeast (354
days), Southern Great Plains (331
days), and Southwest (331 days).
The smallest increases in days
occurred in the Northern Great
Plains (198 days) and Midwest
(199 days).
Percent change:
• • o • #
Change in number of growing degree days in the contiguous 48 states from 1948-2020.
Lake Superior
Seasonality and Climate Change: A Review of Observed Evidence in the United States
-------
SEASON TIMING
Key Point: The timing of peak snowpack
shifted earlier by an average of 9 days
from 1982-2020 in the West.
Peak snowpack timing change (days):
~ ~ ~ ~ * * A a A ~
<-20 -15 -10 -5 0 0 5 10 15 >20
to-20 to-15 to-10 to-5 to 5 to 10 to 15 to 20
Trends in the date when snowpack reaches its deepest level across the western United States
Blue circles represent a shift to later timing; red circles represent a shift to earlier timing.
fW'
w w

3B
ifHn
~ • ~ ,•
. / T
An ~ ~ H
T °

/
f*T
~
*

o
~

Key Point: In parts of the country with
substantial snowmelt, winter-spring
runoff is happening at least five days
earlier than in the mid-20th century
at most gauges. The largest changes
have occurred in the Pacific Northwest
and Northeast.
~ ~ * ° » A ~
More than 5 to 10 2 to 5 2 days 2 to 5 5 to 10 More than
10 days days days earlier to days days 10 days
earlier earlier earlier 2 days later later later later
Changes in the timing of annual high winter-spring flow carried by rivers and streams from
1940-2018. Trends are based on the winter-spring center of volume, which is the date when half of
the total January 1-July 31 streamflow (in the West) or half of the total January 1-May 31 streamfiow
(in the East) has passed by each streamflow gauge, reflecting the timing of spring snowmelt.
Seasonality and Climate Change: A Review of Observed Evidence in the United States
-------
SEASON TIMING
January 1
February 1
1840 1860 1880 1900
1920 1940
Year
I960 1980 2000 2020
Detroit Lake
Lake George
Lake Monona
Mirror Lake
Shell Lake
Geneva Lake
LakeMendota
Lake Superior
Otsego Lake

at Bayfield

April 1
June 1
— Cobbosseecontee Lake
— LakeMendota
Mirror Lake
— Damariscotta Lake
— Lake Monona
Moosehead Lake
Detroit Lake
LakeOsakis
— Otsego Lake
— Geneva Lake
— Lake Superior
— SebagoLake
— Lake George
at Bayfield
— Shell Lake
Date of first freeze for nine U.S. lakes and date of ice thawing and breakup for 14 U.S. lakes. The
data are available from as early as 1840 to 2016, depending on the lake, and have been smoothed
using a nine-year moving average.
Key Point: Freeze dates have shifted later at a rate of roughly half a day to one-and-a-half days per
decade. Thaw dates for most of these lakes show a trend toward earlier ice breakup in the spring.
Spring thaw dates have grown earlier by up to 24 days in the past 111 years.
« 15
>>
-U
01
1S
k-
Q)
>
<0
o
(N
00
a\
>
-------
SEASON TIMING
Key Point: Leaf and bloom events
are happening earlier throughout the
North and West but later in much of
the South. This observation is generally
consistent with regional differences in
temperature change.
Change in first leaf date:
Morethan -4to -1 to Within 1 to 4to Morethan
-8 days -8 days -4 days 1 day 4 days 8 days 8 days
< ~
Earlier Later
First leaf date trends in lilac and honeysuckle at weather stations across the contiguous 48 states.
This map compares the average first leaf date for two 10-year periods (1951-1960 and 2011-2020).
March 1
National Cherry Blossom Festival
March 11
_§ March 21 IN A I I &
j:::
April 20
April 30
1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020
Year
Peak bloom date each year for the main type of cherry tree around the Tidal Basin in Washington,
DC. The peak bloom date occurs when 70 percent of the blossoms are in full bloom. The shaded
band shows the timing of the annual National Cherry Blossom Festival.
Key Point: Peak bloom
date for the cherry trees is
occurring earlier than it did
in the past. Since 1921, peak
bloom dates have shifted
earlier by approximately six
days. The average peak bloom
date is April 4th.
Seasonality and Climate Change: A Review of Observed Evidence in the United States
-------
SEASONAL VARIABILITY
6,000
5,000
4'000
>»
~U
S 3,000
U)
0)
Q 2,000
Heating degree days (colder than 65°F)
Cooling degree days (warmer than 65°F)
1,000
Key Point: Since around
1980, the number of heating
degree days has decreased
and the number of cooling
degree days has increased
relative to the 20th century
average. The recent increase
in cooling days is driven by
more frequent days above
65°F and more frequent
extreme high temperatures.
1890 1910
1930
1950
Year
1970
1990
2010
2030
Heating and cooling degree days in the United States from 189b to 2020.

J
• * '"/'•of
.* .
• &
U H •
Rate of change (percent per year):
• • • •
-0.9
to-1.2
-0.6 -0.3 -0.1 -0.1 0.1 0.3 0.6
to -0.9 to-0.6 to-0.3 to 0.1 to 0.3 to 0.6 to 0.9
0.9
to 1.2
Key Point: Total snowfall has
decreased in many parts of
the country since widespread
observations became available in
1930, with 57 percent of stations
showing a decline. Among all of the
stations shown, the average change
is a decrease of 0.19 percent
per year.
Average rate of change in total snowfall from 1930 to 2007 at 419 weather stations in the
contiguous 48 states. Blue circles represent increased snowfall; red circles represent a decrease.
Seasonality and Climate Change: A Review of Observed Evidence in the United States
-------
SEASONAL VARIABILITY
s C
.is o
u Irt
£ *
t; el
ai ¦_
* s.
s 5
V) +¦»
Of <0
;i
£ 3
£ -
3
1/1
1,600
1,400
1,200
1,000
800
600
400
200

\\ J
pH''

\
ft!
\! v
\ X - i-'j
\ \ / rS
\ i v Jfyi /
* I ^ ' \
« 1 V
V
V
'—V

V *

¦ Summer electricity
use per capita (50 states)
Summer cooling degree
days (contiguous 48 states)
0
1970
1,200
1,050
900
750
600
450
300
150
0
>-
IB
TS
31
-------
SEASONAL VARIABILITY
a

01
73
/Cey Point: The amount
of natural gas used by
1,500 the average American at
home during the winter
has decreased since 1973.
a
1,000 f
Si
c
g
500
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025
Year
Residential winter natural gas use per capita and winter heating degree days in the United
States, 1974-2020. The black line shows average winter natural gas use per capita, and it represents
all 50 states plus Washington, DC. For reference, the blue line shows the average number of
heating degree days for the same months across the contiguous 48 states plus DC.
1996
2018
The distribution of Lyme disease cases in the United States, which is reported to CDC in 1996 and
2018. Each dot represents an individual case placed according to the patients' county of residence,
which may be different from the county of exposure. These maps focus on the parts of the U.S.
where Lyme disease is most common.
Key Point: The incidence of Lyme disease has approximately doubled since 1991.Driven by multiple
factors, the number and distribution of reported cases of Lyme disease have increased over time.
Note that because these data only include diagnosed cases of Lyme disease, the actual incidence
is likely higher.
Seasonality and Climate Change: A Review of Observed Evidence in the United States
-------
SEASONAL VARIABILITY
on °
oS*o° o

Key Point: Nearly 80 percent of
the stations across the contiguous
48 states have experienced a
decrease in the proportion of
precipitation falling as snow.
Percent change:
•
•
•
•
•
•
•
•
•
•
•
<-40
-30
-20
-10
-2
-2
2
10
20
30
>40

to -40
to-30
to-20
to-10
to 2
to 10
to 20
to 30
to 40