https://orcid.org/0000-0001-9103-6065
Figures
Abstract
Heat waves are the last extreme weather events without a formal, on the books, definition. Instead, across the U.S. those working on extreme heat event management, forecasting, and planning are using differing definitions in their work. With such differing definitions being used there are widespread impacts including some to human and environmental health, natural resource management, and long-term emergency management planning. For instance, when should heat advisories for vulnerable populations be released when an event impacts a region using multiple definitions? There are concrete and justifiable reasons for the lack of a formal heat wave definition including, at its simplest, differences in what temperature is extreme enough, compared to the region’s climatological regimens, to be deemed as an extreme heat event or heat wave. This study looks for patterns and commonalities in emergency managers and climatologists, those most commonly addressing or planning for such events, definition of heat wave events through a review of the literature and widespread survey across the United States. Through a short 11-questions survey and subsequent text mining, we find widespread variability in the common heat wave definitions but a consistent pattern of core key term usage including aspects of heat duration, extreme temperature, and humidity. However, we also see little to no usage of non-climatological variables such as exposure, vulnerability, population, and land cover/land use.
Citation: Bunting EL, Tolmanov V, Keellings D (2024) What is a heat wave: A survey and literature synthesis of heat wave definitions across the United States. PLOS Clim 3(9): e0000468. https://doi.org/10.1371/journal.pclm.0000468
Editor: Teodoro Georgiadis, Institute for BioEconomy CNR, ITALY
Received: December 14, 2023; Accepted: July 13, 2024; Published: September 5, 2024
Copyright: © 2024 Bunting et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data needed to recreate this article is provided in the supplemental materials associated with the article. Identifying data will not be released including the survey takers names or contact information, but these data are not essential for study recreation. This project was deemed exempt from full IRB review. Any questions related to the data, or the article can be addressed by emailing the authors Erin Bunting (ebunting@msu.edu) or David Keellings (djkeellings@ufl.edu). If you have any data requests or additional questions, and the authors are not available, please feel free to contact MSU RS&GIS (info@rsgis.msu.edu). This group has prior knowledge of the project and a copy of project and IRB documents on a secure server. If you have any questions or concerns regarding the survey, contact please contact the Michigan State University Human Research Protection Program (HRPP) office at 517-355-2180 or via email at IRB@msu.edu. Please use the following MSU Study ID: STUDY00004016 if you communicate with MSU HRPP.
Funding: This research was funded by a grant from the National Science Foundation GSS program (Award #2203235). Multiple people and organizations made this publication possible and the authors wish to express their gratitude, especially to Dr. Laura Myers and Jacob Reed at the University of Alabama, Dan Wanyama, and the Staff of Remote Sensing and GIS Research and Outreach Services (RS&GIS) at Michigan State University. Additionally, we thank the reviewers for their time and effort put into manuscript review.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Extreme weather and climate events have been impacting human and natural landscape since the beginning of time. However, with changing climate patterns we are seeing greater impacts of these events globally. As stated in the most recent IPCC report it is an “established fact” that human induced change has resulted in “an increased frequency and intensity of some weather and climate extremes since pre-industrial times” [1]. With such changing patterns it is important to look at not just the impacts and patterns of such events but the premise of the event definition itself. Extreme weather and climate events are broadly defined as severe weather or climate conditions that induce devastating impacts to the human and natural landscapes. While such events can be weather-related (short in duration), or climate related (long in duration) there is a basic understanding that the event is defined as atypical and beyond the normal. Almost all these extreme events (e.g., hurricanes, tornadoes, blizzards) have standard definitions related to what triggers the event, how they are measured, and their severity classification. For instance, with hurricanes, the Saffir Simpson class is used to define the pressure and wind speeds associated [2]. With tornadoes, the Enhanced Fujita Scale (EF) uses derived engineering wind estimates in assessing strength and resulting damage [3].
While almost all climatological events have standard definitions one of the deadliest does not, heat waves. Over the past decade, heat waves of varying durations and intensities have impacted much of the globe. For instance, the 1995 heat wave in the central United States resulted in more than 1000 deaths [4, 5]. Further, across France, the large 2003 European heat wave resulted in excess mortality of approximately 15,000 individuals, up 60% from normal mortality patterns [6, 7]. In line with the IPCC remarks on trends of extreme events there has been an increased frequency of heat waves not just across the US and France, but across Europe, China, Australia as well [1]. Additionally, it has been projected that heat waves will be more frequent, intense, and longer-lasting into the future [1, 8–10], thus raising concerns on planning for such events.
There is no worldwide consensus on a heat wave definition though its usually thought simply as an extended period of extreme heat. Not grounded in academic or climatological literature, the Merriam dictionary defines heat waves as “a period of unusually hot weather”. Additionally, in many published articles, simple definitions based purely on temperature are utilized, including those as simple as “an extreme heat event is defined when the temperature exceeds a given threshold with an appropriate spatial extent” [11]. Attempts have been made at standardizing a heat wave definition. For instance, in 1996 Environment Canada provided a more scientifically grounded definition of a heat wave as a period of more than three consecutive days of maximum temperatures at or above 32 degrees Celsius [12]. Additionally, governmental groups both in the US and around the world have developed de facto definitions of heat waves including: the US National Weather Service, NOAA, and UK Met Office. These definitions are being developed for the issuance of heat watches and warnings and therefore should be pertinent for use by emergency managers and state climatologists. With this in mind, one would expect high overlap between national government definitions of heat wave and those from managers and state climatologists. The aforementioned governmental groups define heat waves as such:
- US National Weather Service: A period of abnormally hot weather generally lasting more than 2 days. Heat waves can occur with or without high humidity.
- NOAA: A period of abnormally and uncomfortably hot and unusually humid weather. Typically, a heat wave lasts two or more days.
- UK Met Office: An extended period of hot weather relative to the expected conditions of the areas at that time of year, which may be accompanied by high humidity.
- World Meteorological Organization (WMO): A period where local excess heat accumulates over a sequence of unusually hot days and nights.
What is lacking from these definitions is consistency regarding temperature thresholds, metrics, durations, or number of days used to define such events [13]. Further, even using these definitions, adverse heat impacts on human health have been documented at lesser extreme temperatures and durations [12, 14, 15]. It’s important to note that the use of various heat wave definitions results in temporal variability in heat wave classification, inability to compare events and synthesize results across regions, and inconsistent terminology in the literature. Many of these challenges can be overcome by finding some consistency across regions and definitions.
Why it is difficult to develop a consistent heat wave definition is fairly obvious, heat waves differ in their intensity (magnitude), extent, duration, and scope of impact [16]. Numerous studies have used different thresholds of mean or maximum temperature [17], percentiles of maximum temperature, heat indices, or even combinations of thresholds [18, 19]. The common variables of these studies being the use of intensity and duration factors. [20] looked at 45 definitions of heat waves, combining 5 temperature thresholds, three temperature indicators (daily mean temperature, minimum temperature, and maximum temperature), 5 percentile metrics (90th, 92.5th, 95th, 97.5th, and 99th), and multiple event duration lengths (2, 3, and 4 days) to assess how different definitions align with mortality patterns. Overall, [20] found the best model fit, and therefore best heat wave definition, using daily mean temperature in the 99th percentile in combination with a 3-day event duration. Similarly, two studies, one in the US and another from West Africa, found the best fit model was produced by a heat wave definition using both minimum and maximum temperatures in the 90th percentiles with an event duration of 3 days [21]. Drawing from these, and other, studies we can see a heat wave definition needs to include factors of: (1) intensity or magnitude: based on a tested index or temperature threshold, (2) duration: defining the persistent of an event to be a heat wave, (3) extent: geographic areas impacted and measures of exposure, and possibly (4) severity [22].
There are no doubt other factors contributing to heat wave impacts such as frequency, timing, event size, and population density. Incorporation of these factors into a unifying heat wave definition is fraught with complications. For instance, a higher frequency of hot and humid conditions do not necessarily result in a heat wave and severe heat wave impacts [12]. Similarly, regions experiencing more hot and humid summer conditions already have physiological, behavioral, and infrastructure adaptations to extreme heat, likely reducing the harmful effects [12]. However, with such heat wave components (intensity, duration, and extent), we can begin to understand the social, cultural, and physical impacts of extreme heat events. For instance, with such a holistic definition we can begin to develop temporary modification to lifestyles to minimize heat stress exposure and impacts. More clearly such a definition enables evasive action and management practice to be developed.
Heat waves can be defined in several ways, through absolute and relative approaches. With an absolute heat wave definition an exact event duration threshold would be set in tandem with a pre-determined temperature and/or heat stress index level [12]. Whereas, if a relative heat wave definition were developed it would have to take into consideration acclimatization to weather, exposure, and human dimensions in addition to region specific climate trends. Most published studies on heat wave definitions look to model or develop the core metrics, threshold, and durations to define event occurrence. In this study we ask those on the ground for insight into heat wave event definition. Through a simple survey conducted across the United States in 2020 we look at what definitions each respondent is currently using, what variables they see as critical, impacts of differing definitions, and how relative definition approaches are used in their work. Overall, it was hypothesized that (1) emergency managers would have a different perspective on heat wave definition, especially as it relates to human exposure and acclimatation, (2) differing local to regional climate trends across the country would result in different definitions north to south across the U.S., and (3) only atmospheric variables would be considered in respondents’ heat wave definitions.
2. Data and methods
2.a. Survey development
The survey was designed to see how those involved in heat wave forecasting and management define such events and what factors contribute to their definition. The 11-question survey also included space for respondents to provide additional text and information to clarify their definitions and provide other important details. The survey was developed, tested, and IRB approved at Michigan State University and the University of Alabama. Respondents, none of whom were minors, were informed of survey privacy and provided consent language prior to taken the survey. All respondents provided written consent at the onset of the survey. The survey was developed within Qualtrics and consisted of 11 questions (Fig 1).
The survey commences with questions of occupation, geographic extent of work, and zip code of residence (Fig 1). Question 4–11 are, using free response, asking how, in their professional capacity they, define heat wave events. Additionally, we ask: what variables are associated with the definition (in case this is not mentioned in the free response), what non-atmospheric variables are included, and looks to understand how event duration, size, and time of year factor into their definition.
The survey was open for approximately five months. Distribution of the survey occurred through emails to: (1) state climatologists listserv, (2) emergency management associations, (3) a contact list for all state-recognized emergency managers, and (4) the American Meteorological Association listserv. Overall, the survey was disseminated through associations and email groups related to the parties of interest.
2.b. Data processing and analysis
Once the survey period was completed the Qualtrics survey was closed, and the data were downloaded locally. To clean the data, we first looked at the completeness of responses. Overall, 137 individuals fully completed the survey and 25 partially completed the survey. It was decided if a respondent answered at least half of questions 4–11 (see Fig 1) that these would still be included in the analysis. Other responses were removed from the analysis.
Data analysis occurred in several ways. Text mining and pattern analysis were conducted within R [23]. Initially, analysis began with simple frequency counts of factors such as geographic extent of work, field of work, etc. Next, for questions four and five, text mining was conducted using several R packages including SnowballC [24]. First, by question, the responses were merged into a corporal collection of phrases containing natural text. From there the tm_map function was used to remove symbols, number, punctuation, and common words (i.e., cause, the, and, is, have, are, was, be, of) from each survey response entry. The resulting text is not in sentence format but rather the key words within each response.
After the data were cleaned for text mining, analysis began with simple frequency counts. Using this simple statistic, the common terms or words utilized by the respondents were tallied by question. Terms or words used more than twice were preserved in the analysis. In addition to the frequency analysis, word clouds were constructed using the wordcloud package in R [14].
The word cloud represents the extremes of thought, terminology, and definition of heat waves. The word clouds also highlight the frequency in term utilization by the respondents. For the remaining questions simple summary statistics were completed in R.
3. Results
3.a. Profile of survey respondents
We see large diversity in the response pool, both spatially and across profession. Overall,162 individuals took part in this project with 137 fully completing the survey. This corresponds to an 84.55% completion rate. The 162 responses came from across the US and across a wide variety of subfields related to climatology, emergency management, and meteorology. Spatially, responses were collected from 43 of the 50 states with Hawaii, South Carolina, Massachusetts, Rhode Islands, Connecticut, Virginia, and West Virginia being the exceptions. While there are no direct responses from those states, they are partially represented by those that work regionally, nationally, or at the global scale. While there are survey responses across the country there is a slight skew to the southern portion of the United States. For those respondents that work county to statewide we see the most responses from Alabama, Mississippi, Florida, Arizona, and Texas (Fig 2). However, there are responses at this scale spread evenly across the U.S.
Dots represents those that work at the county level. Colored polygons represent those that work at the state level and the number of people that responded to the survey that work at the state level. Numbers represent those that responded that worked at another scale of geometry (e.g., city, community, multiple counties, townships, etc.). Not represented are those that worked globally (n = 8), nationwide (n = 4), and those that work sub-county level (city or community). Map created in ArcGIS using survey data and a states shapefile from the US Census Bureau (https://www.census.gov/geographies/mapping-files/time-series/geo/carto-boundary-file.html).
The professional profile of those that completed the survey is diverse and does not skew to one group of individuals or those with certain specialties. Eighty-seven (53.7%) of the completed surveys came from those that identified forecasting as their occupation. Of those 62 worked across multiple counties in a single state, 11 at the state level, 7 at the regional (multiple states or parts of multiple states), 3 at the county level, 2 at an “other level” (city to local level), 1 globally, and 1 at the community level. After forecasting the next highest group of responses came from emergency managers. Overall, 46 (28.2%) emergency managers from across the US completed the survey. Of those emergency managers 32 worked at the county level, 2 worked across multiple counties, 1 worked at the regional scale (across multiple states), 1 worked at the national level, 1 worked at the community level, and the remaining 9 worked at other scales (e.g., tribal lands, city, ecoregion, etc.). Lastly, 29 individuals identified their career as “other”, including: researchers, retired state climatologists, climatologists, land managers, and those in academia. From this group the majority worked at the state to global scale.
3.b Defining heat waves
Text mining of survey responses showed interesting keyword usage patterns (Fig 3). For the question “What is your definition of the climatological term ‘heat wave’”, heat was not the most used word, instead days was mentioned 71 times by respondents. Answers related to this term included definitions like: “Several days of 95+ degrees”, “A period of multiple days beyond normal temperatures”, and “At least 3 consecutive days of high maximum temperatures”. After days, the terms heat (n = 68), period (n = 65), normal (n = 47), and high (n = 42) were the other common terms (Fig 3). Overall, this word usage pattern highlights a highly important and common thought pattern in defining heat waves, such extreme events have a duration aspect that needs to be considered and defined. Heat waves are multi-day extreme events.
Response trend word clouds and frequency counts for the questions: (A) “What is your definition of the climatological term heat wave?”, and (B) “What atmospheric variables are part of your definition of a heat wave?”, and (C) “What other atmospheric variables are part of your definition of a heat wave”.
When asked, “What atmospheric variables are part of your definition of a heat wave”, the most selected term was maximum temperature (n = 129) followed by heat index (n = 85), minimum temperature (n = 75), and humidity (n = 74). These four terms were far more common than the next terms of the list, average temperature (n = 45) and other (n = 25). Respondents were given a list of possible terms for this question and asked to select all that apply to their definition of heat waves. The terms included include maximum temperature, minimum temperature, average temperature, humidity, heat index, and other. This word usage pattern highlights that extreme terms are central to the definition of heat waves.
Lastly, respondents were asked “What other atmospheric variables are part of your definition of a heat wave?” This question was asked so respondents have free response, instead of choosing from a bank of options as with the previous questions, the climatological variables that they use in management, forecasting, and planning. It is important to note most respondents did not list any additional atmospheric variables as part of their heat wave temperature. Of those that did the most common response was wet bulb globe temperature (WBGT), wind, cloud cover, and insolation. A few respondents listed, repetitively, humidity and heat index, in response to this question. Overall, this word usage pattern highlights the strong pattern of traditional climatological variables in the definition of heat waves, rather than characteristics of land, demographics, human health, or exposure.
3.c Beyond climatological terms, other factors to considered in defining heat waves
Beyond traditional climatological terms we asked respondent “Does your definition of a heat wave include non-atmospheric variables?” Only 27.5% of respondents replied in the affirmative that they did include non-atmospheric variables in their definition of a heatwave (Fig 4). The common non-atmospheric variables included in the heat wave definition were grouped into categories of: (1) impact on humans, (2) seasonal variation, (3) physical variables, and (4) other. The highest percentage of responses were within the impact to humans categories with 35.3%. Common non-atmospheric variables listed by respondents included: soil moisture, land type, duration, percent impervious surfaces, and crop stress. Most respondents that included non-atmospheric variables in their definition of heat waves were forecasters (n = 25), with only 5 emergency managers and those identified as “other” in their career including non-atmospheric variables. Geographically, those that included non-atmospheric variables mostly worked at the state to multi county scale (n = 22). Others that included non-atmospheric variables worked at varying geographic extents including county scale (n = 7), regional scale (n = 3), and global scale (n = 2). Spatially, that 27.5%, where not clusters in one portion of the US. Those that included non-atmospheric variables spanned from Arizona to Vermont.
Response trends to the questions: (A) “Does your definition of a heat wave include non-atmospheric variables?” and (B) “If yes, what other variables do you include?”.
Next, respondents were asked if their heat wave definition included a threshold that must be surpassed to be considered a heat wave and what threshold measurement they used. More than 56% of respondents said yes, their definition included a threshold that must be crossed. Of these 18 were emergency managers, 33 were forecasters, and 16 listed other as their career. As such, of the respondents that completed the survey approximately 38% of forecasters, 39% of emergency managers, and 55% of others included a threshold in their heat wave definition. Spatially, those that included a threshold worked across all geographic extents from community / local to global, though slightly more worked at the statewide or multicounty scale.
It was thought that the main threshold both emergency managers and forecasters would use in their definition would be related to temperature. Many of the provided definitions stated something like:
“Temperature above 95 degrees Fahrenheit for an extended period of time”
Or
“A persistent anomaly in daily surface temperature usually many days about the 98th percentile.”
Instead, survey results show that only 45.7% of respondents reported using temperature thresholds in their heat wave definitions, followed by heat index (27.9%), heat duration (23.5%), and heat risk (Fig 5). Spatially, if we look at the common temperature threshold mentioned by survey respondents there is an interesting dynamic playing out (Table 1). Those surveyed from the Northeast, Southwest, and Central regions of the United States, as defined by the NOAA climate regions, all listed a temperature threshold of 90°F with little variability. The lowest temperature threshold mentioned was 80°F and it was from a respondent in the south region. Whereas the highest threshold reported was 105°F, occurring in both the East North Central and South regions. Overall, a north to south temperature threshold gradient was expected, as the more southerly regions are subject to climatically higher mean temperatures, but this was not seen in the survey responses.
Response trends to the questions: (A)” Does your definition of a heat wave include a threshold that must be crossed in order to be considered a heat wave?”, and (B) “what is the threshold measurement that you use?”.
Fig 5B shows that numerous respondents suggest a threshold in their heat wave definition based on duration. During the survey respondents were specifically asked “Does your definition of a heat wave include a minimum duration”. This question was asked, in tandem with the aforementioned in case a respondent did not feel duration was a threshold measure. Over 74% of respondents said yes that event duration was a part of their heat wave definition. Of these 25 were emergency managers (54.3% of EM total population), 56 where forecasters (64.4% of F total population, and 19 listed their career as other. Importantly, the most noted durations for events were 2 and 3 days (Fig 6). Overall, 22.2% or respondents mentioned 2 days as the minimum duration for a heat wave to be defined and 56.6% listed 3 days as the minimum duration. The response pattern of did not vary by geographic extent of work or the region respondents work in.
Response trends to the questions: (A) “Does your definition of a heat wave include a minimum duration?”, and (B)” If yes, how many days?”.
At this point, respondents had solely been asked about common aspects of heat wave definitions such as duration, intensity, etc. The next question of the survey asked, “Is size of area affected part of your definition of a heat wave?” Only 21.8% of respondents considered event size in their heat wave definition, equating to just 29 respondents. Of these there was an even split between emergency managers and forecasters, the geographic extent of their work was mostly single or multiple counties, and almost all included duration in their heat wave definition.
The majority of those that responded in the affirmative said that a heat wave had to have a duration of 2 or 3 days to be defined as such an event. The range of durations suggested by those surveyed was as short as 1 day (3 respondents) and as long as 10 days (1 respondent) (Fig 7).
Response trends to the questions: (A) “Is size of area affected part of your definition of a heat wave?”, and (B) “If yes, how is size incorporated into your definition of a heatwave?”.
If a heat wave only impacts a small area, does it matter in emergency management and forecasting as much as large events? That question was the impetus for the next survey question, “is size of area affected part of your definition of a heat wave.” The vast majority (78.2%) of people said no, size does not matter. Those that answered yes to this question were almost all forecasters (except 2 out of 14). Those that answered in the affirmative had a clear theme, larger events impact more people compared to isolated events and that is why event size matters. With this theme several people mentioned that an event needs to be state-wide or span multiple counties (44.8% of the previous questions affirmative answers).
The last two questions of the survey look to see how respondents think about defining heat waves across space and time. Respondents were asked “Is your definition of a heat wave dependent on time of year?” This question was almost a 50/50 split in terms of percentages with 53% saying the definition is not time of year dependent and 47% saying it is (Fig 8). For those that answered in the affirmative they were asked how their definition varies through the year. There were a wide range of responses to this follow up question from those mentioning seasonal threshold values to others saying only summer matters as it’s the warmest season. Overall, the consensus of those surveyed is well represented by one respondent’s remakes, “I generally only call things "heat waves" when it is “hot” outside. I wouldn’t call a winter stretch of warmer than normal weather a "heat wave" unless it was drastically warmer than normal.” There is a consistent trend in these types of answers with others saying outright “heat waves have only occurred climatologically in our late spring to early fall months, when it’s hottest”. However, there were a small group (labeled statistical in Fig 8) that again link back to thresholds and say, in their opinion, that heat waves can happen if the temperatures exceed the 95th percentile for that given time frames normal.
Response trends to the questions: (A)” Is your definition of a heat wave dependent on time of year?”, and (B) “If yes, what time of year?”.
Lastly, respondents were asked if their definition of a heat wave varies across space. This question was asked to see if one core definition of heat waves cannot be developed and instead a regional definition would be needed. Overall, only 36.3% responded in the affirmative whereas 63.7% said their heat wave definition did not vary across space (Fig 9). Those answering in the affirmative, that their definition varied across space, were then asked how with the options of: (A) regional differences, (B) type of territory, (C) statistically, (D) varies by the gridded data, and (E) other. Overall, 42.9% of respondents that thought heat wave definitions should vary across space thought there were regional differences. When asked to explain their answer respondents had a variety of answers and mentioned topic such as: “definitions can be different across climate regions because citizens are acclimated to different levels” and “In my work the western high plains should have a higher threshold for defining heat waves than the eastern portion of the Southern Plains”, and “Since humans can become acclimated to "normal" conditions, I would vary the definition based upon a certain amount above climatology”. Next, 22.5% of respondent said heat wave definitions should vary because of type of territory. Those that responded in this manner mentioned: “amount of vegetation vs bare ground is important” and “population density and impervious surfaces need to be considered to account for urban heat island effects”. Third, 22.5% said statistically heat wave definitions should vary spatially. These respondents provided further understanding of their response with comments such as: “Heat wave might be the top 0.1% of high temperatures of all time for each location” and “temperature percentile varies by location”. Lastly, only 6.1% of respondents said their definition varies over space because of gridded data. Specifically, those that answered in this manner added comments like: “heat risk is calculated on a spatial grid by entire forecast area” and “gridded temperature data is 2.5 km resolution—so the values for defining the heat wave vary spatially even if the definition remains the same.” Overall, these responses highlight a need to better understand duration, thresholds, and the underlaying population exposure.
Response trends to the questions: (A) “Does your definition of a heat wave vary across space?”, and (B) “How does it vary spatially?”.
4. Discussion and conclusions
Heat waves are regarded by the U.S. National Weather Service as a major, in fact leading, cause of weather-related fatalities in the U.S. in most years [18]. Closing the gap on a formal heat wave definition is important because of such direct affect that these extreme events can induce. A formal definition is important to multiple subfields, beyond emergency management and forecasting, as in the literature it has been consistently noted that the direct adverse effects of heat waves include (1) increased demands for water and electricity [25–28], (2) reduced productivity and overall labor efficiency [25], (3) drought and overall crop stress or failure, and (4) human health, including cardiovascular and respiratory damage in addition to death. Further, in the literature heat-related studies have been published related to land management, policy development, vulnerability/exposure regarding human health, urban development, emergency planning, and ecosystem health.
Overall, our population is vulnerable to heat waves and therefore we need a better understanding of these extreme events. For instance, the impacts of heat waves on human health are widely documented regarding mortality and morbidity [29, 30]. That said, heat wave vulnerability is unequal and unevenly distributed across both human and natural landscapes. The elderly, those residing in nursing homes, and the chronically ill are readily identifiable as susceptible subgroups, high at risk for extreme heat impacts, and studies have criticized the lack of effective heat management for such populations [31–35]. These populations are not the only ones impacted by heat waves and the way such events are managed. Children have a higher sensitivity, outdoor workers have more extensive exposure, and the homeless are just a few of the other vulnerable populations [36–40].
Heat related deaths occur when a rapid temperature increase outpaces the body’s ability to cool itself, though perspiration and increased blood circulation [41]. There are compounding factors to such mortality risk including high humidity and overall exposure. If we examine the characteristics of heat waves mentioned throughout this study and our survey (duration, intensity, timing, size) its known that these factors negatively impact public health by increasing the risk of heat-related mortality [41]. For instance, long duration heat events increase exposures, elevate even the daily minimum temperature, and limit the body’s ability to recover [17, 27]. Additionally, large-scale events are more likely to expose broader human populations to such extremes, increasing the population vulnerable to such an event. Lastly, the timing of an event can have multiple implications including (1) early extreme events can result in large populations of people unprepared and (2) events during peak summer months can be assumed to be more intense in nature. Beyond health impacts, the timing of such events can alter soil available moisture, impacting plant phenology and productivity. With such changes in soil moisture there are not only connections to widespread drought but also to impacts of cropping and overall harvest quality and quantity, again linking back to human impacts. Overall, in the US, the Center for Disease Control (CDC) estimates that heat related deaths average over 1000 per year [42]. However, the impacts of extreme heat stretch far beyond the health ramifications and therefore needs to be studied and defined more holistically.
How can a formal unified heat wave definition impact emergency management? Simply, with the development of a formal definition and associated research emergency managers and forecasters can know what to expect, outcome wise and in knowing the vulnerable populations, and plan for during extreme heat events. Much research has gone into analyzing heat wave definitions to see how they align with differing outcome patterns, particularly human health impacts. For instance, Knowlton et al, analyzed heat wave induced hospitalizations and emergency room visits resulting from the 2006 California extreme event and found that a definition based on a higher maximum temperature threshold was associated with a greater relative risk for hospital admission [43, 44]. Other studies have tested heat wave definition to decide the ideal scenario for the opening of cooling centers and to consider moving vulnerable people to safer locations [45, 46]. Beyond hospitalization and evacuation with a formal heat wave definition managers can understand, through research of trends, what to expect for certain mortality risks. For example, Dong et al. 2016 found that a heat wave definition using the 93rd percentile of maximum temperature and 5 day event duration was the best way to understand the trends between extreme heat and cardiovascular mortality [47]. These examples, explicitly related to health outcomes and exposure due to extreme heat are just a few ways in which emergency management could use such information and data.
We live in a warming climate, and with such changing regimes it’s projected that the duration, intensity, and frequency of heat waves will increase [48]. From the literature and survey, we can surmise a few key aspects of a formal heat wave definition. Factors that need to be considered in the formal definition include duration, intensity (a climatological threshold), and exposure. Survey results show that the majority respondents report a heat wave can be defined as an event that lasts at least 3 days. The mean duration listed by survey respondents was 3.14, the mode was 3 days, and the median was 3 days as well. Overall, the durations listed by respondents ranged from 1–10 days. Regardless of occupation and geographic extent of work most respondents felt a threshold value needed to be crossed for a heat wave to formally be defined. Both percentiles and numeric thresholds were suggested. Additionally, a large group of respondents felt that maximum temperature was the key metric to develop such a threshold. For those that wanted to include percentiles in their heat wave definition the 95th percentile was the most noted. In terms of absolute temperatures there was a large range of possible thresholds suggested, 90–105 degrees. The mean temperature threshold suggested by respondents was 94.6 degrees. Interestingly, there were no survey response trends suggesting for spatial variation in heat wave definitions. In terms of time of year, most felt that heat waves would only be defined in the hottest months and are not solely associated with any monthly or seasonal temperatures beyond the normal. One of the most important non-climatological aspects brought up by survey respondents was exposure and associated population density. Such survey results do not solidify a formal heat wave definition, but they do show extremely similar thought processes across different occupations and regions of the US. With such similarities in defining heat waves there is no doubt similar patterns of use for such data in emergency management and forecasting, and perhaps for a future unified definition of heat waves.
Supporting information
S1 Data. Deidentified survey data collected including the full IRB information and question in the header of the data file.
https://doi.org/10.1371/journal.pclm.0000468.s001
(XLSX)
Acknowledgments
Multiple people and organizations made this publication possible, and the authors wish to express their gratitude. The authors wish to especially thank Dr. Laura Myers and Jacob Reed at the University of Alabama, Dan Wanyama, and the staff of Remote Sensing and GIS Research and Outreach Services (RS&GIS) at Michigan State University. Additionally, we thank the reviewers for their time and effort put into manuscript review.
References
- 1.
Lee H, Calvin K, Dasgupta D, Krinner G, Mukherji A, Thorne P, et al. IPCC, 2023: Climate Change 2023: Synthesis Report, Summary for Policymakers. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, Lee H. and Romero J. (eds.)]. IPCC, Geneva, Switzerland. 2023 URL: https://mural.maynoothuniversity.ie/17886/
- 2. Dolan R, Davis RE. An intensity scale for Atlantic coast northeast storms. Journal of coastal research. 1992; 840–853.
- 3. Doswell CA III, Brooks HE, Dotzek N. On the implementation of the enhanced Fujita scale in the USA. Atmospheric Research. 2009;93: 554–563.
- 4.
Klinenberg E. Heat wave: A social autopsy of disaster in Chicago. University of Chicago press; 2015.
- 5. Palecki MA, Changnon SA, Kunkel KE. The nature and impacts of the July 1999 heat wave in the midwestern United States: learning from the lessons of 1995. Bulletin of the American Meteorological Society. 2001;82: 1353–1368.
- 6. Fouillet A, Rey G, Wagner V, Laaidi K, Empereur-Bissonnet P, Le Tertre A, et al. Has the impact of heat waves on mortality changed in France since the European heat wave of summer 2003? A study of the 2006 heat wave. International journal of epidemiology. 2008;37: 309–317. pmid:18194962
- 7. Pascal M, Wagner V, Le Tertre A, Laaidi K, Honoré C, Bénichou F, et al. Definition of temperature thresholds: the example of the French heat wave warning system. Int J Biometeorol. 2013;57: 21–29. pmid:22361805
- 8. Diffenbaugh NS, Ashfaq M. Intensification of hot extremes in the United States. Geophysical Research Letters. 2010;37.
- 9. Barriopedro D, Fischer EM, Luterbacher J, Trigo RM, García-Herrera R. The hot summer of 2010: redrawing the temperature record map of Europe. Science. 2011;332: 220–224. pmid:21415316
- 10. Meehl GA, Tebaldi C. More intense, more frequent, and longer lasting heat waves in the 21st century. Science. 2004;305: 994–997. pmid:15310900
- 11. Stefanon M, D’Andrea F, Drobinski P. Heatwave classification over Europe and the Mediterranean region. Environmental Research Letters. 2012;7: 014023.
- 12. Smoyer-Tomic KE, Kuhn R, Hudson A. Heat wave hazards: an overview of heat wave impacts in Canada. Natural hazards. 2003;28: 465–486.
- 13. Perkins SE, Alexander LV. On the measurement of heat waves. Journal of climate. 2013;26: 4500–4517.
- 14. Kalkstein LS, Davis RE. Weather and human mortality: an evaluation of demographic and interregional responses in the United States. Annals of the Association of American Geographers. 1989;79: 44–64.
- 15. Smoyer KE, Rainham DG, Hewko JN. Heat-stress-related mortality in five cities in Southern Ontario: 1980–1996. International Journal of Biometeorology. 2000;44: 190–197. pmid:11131291
- 16. Anderson GB, Bell ML. Heat waves in the United States: mortality risk during heat waves and effect modification by heat wave characteristics in 43 US communities. Environmental health perspectives. 2011;119: 210–218.
- 17. Hajat S, Armstrong B, Baccini M, Biggeri A, Bisanti L, Russo A, et al. Impact of high temperatures on mortality: is there an added heat wave effect? Epidemiology. 2006; 632–638. pmid:17003686
- 18. Robinson PJ. On the definition of a heat wave. Journal of Applied Meteorology and Climatology. 2001;40: 762–775.
- 19. Weisskopf MG, Anderson HA, Foldy S, Hanrahan LP, Blair K, Török TJ, et al. Heat wave morbidity and mortality, Milwaukee, Wis, 1999 vs 1995: an improved response? American journal of public health. 2002;92: 830–833. pmid:11988455
- 20. Zhang Y, Feng R, Wu R, Zhong P, Tan X, Wu K, et al. Global climate change: impact of heat waves under different definitions on daily mortality in Wuhan, China. Global health research and policy. 2017;2: 1–9.
- 21. Moron V, Oueslati B, Pohl B, Rome S, Janicot S. Trends of mean temperatures and warm extremes in northern tropical Africa (1961–2014) from observed and PPCA-reconstructed time series. Journal of Geophysical Research: Atmospheres. 2016;121: 5298–5319.
- 22. Faye M, Dème A, Diongue AK, Diouf I. Impact of different heat wave definitions on daily mortality in Bandafassi, Senegal. PloS one. 2021;16: e0249199. pmid:33819272
- 23.
R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2023. Available: https://www.R-project.org/
- 24. Bouchet-Valat M, Bouchet-Valat MM. Package ‘SnowballC.’ R package version 07 0. 2020.
- 25. Yin C, Yang F, Wang J, Ye Y. Spatiotemporal distribution and risk assessment of heat waves based on apparent temperature in the one belt and one road region. Remote Sensing. 2020;12: 1174.
- 26. Hatvani-Kovacs G, Belusko M, Pockett J, Boland J. Assessment of Heatwave Impacts. Procedia Engineering. 2016;169: 316–323.
- 27. Anderson BG, Bell ML. Weather-Related Mortality. Epidemiology. 2009;20: 205–213. pmid:19194300
- 28. van Vliet MTH, Yearsley JR, Ludwig F, Vögele S, Lettenmaier DP, Kabat P. Vulnerability of US and European electricity supply to climate change. Nature Clim Change. 2012;2: 676–681.
- 29. Kovats RS, Hajat S. Heat stress and public health: a critical review. Annu Rev Public Health. 2008;29: 41–55. pmid:18031221
- 30. Sheridan SC, Lin S. Assessing variability in the impacts of heat on health outcomes in New York City over time, season, and heat-wave duration. EcoHealth. 2014;11: 512–525. pmid:25223834
- 31.
Brown S, Walker G. Understanding heat wave vulnerability in nursing and residential homes. Comfort in a Lower Carbon Society. Routledge; 2013. pp. 59–68.
- 32. Gupta R, Barnfield L, Gregg M. Overheating in care settings: magnitude, causes, preparedness and remedies. Building Research & Information. 2017;45: 83–101.
- 33. Skinner MW, Yantzi NM, Rosenberg MW. Neither rain nor hail nor sleet nor snow: Provider perspectives on the challenges of weather for home and community care. Social science & medicine. 2009;68: 682–688. pmid:19091451
- 34. Rest K-D, Hirsch P. Supporting urban home health care in daily business and times of disasters. IFAC-PapersOnLine. 2015;48: 686–691.
- 35. Oudin Åström D, Schifano P, Asta F, Lallo A, Michelozzi P, Rocklöv J, et al. The effect of heat waves on mortality in susceptible groups: a cohort study of a mediterranean and a northern European City. Environmental health. 2015;14: 1–8.
- 36. Bethel JW, Harger R. Heat-related illness among Oregon farmworkers. International journal of environmental research and public health. 2014;11: 9273–9285. pmid:25198688
- 37. Vanos JK, Kalkstein LS, Sanford TJ. Detecting synoptic warming trends across the US Midwest and implications to human health and heat-related mortality. International Journal of Climatology. 2015;35: 85–96.
- 38. Nicolay M, Brown LM, Johns R, Ialynytchev A. A study of heat related illness preparedness in homeless veterans. International journal of disaster risk reduction. 2016;18: 72–74.
- 39. Mayrhuber EA-S, Dückers ML, Wallner P, Arnberger A, Allex B, Wiesböck L, et al. Vulnerability to heatwaves and implications for public health interventions–A scoping review. Environmental Research. 2018;166: 42–54. pmid:29859940
- 40. Al-Sayyad AS, Hamadeh RR. The burden of climate-related conditions among laborers at Al-Razi health centre, Bahrain. Journal of the Bahrain Medical Society. 2014;25: 5–8.
- 41. Habeeb D, Vargo J, Stone B. Rising heat wave trends in large US cities. Natural Hazards. 2015;76: 1651–1665.
- 42. Spencer MR, Garnett MF. QuickStats: Percentage distribution of heat-helated deaths,* by age group-National Vital Statistics System, United States, 2018–2020. MMWR: Morbidity and mortality weekly report. 2022;71: 808–808.
- 43. Tong S, Wang XY, Barnett AG. Assessment of heat-related health impacts in Brisbane, Australia: comparison of different heatwave definitions. PloS one. 2010;5: e12155. pmid:20730050
- 44. Knowlton K, Rotkin-Ellman M, King G, Margolis HG, Smith D, Solomon G, et al. The 2006 California heat wave: impacts on hospitalizations and emergency department visits. Environmental health perspectives. 2009;117: 61–67. pmid:19165388
- 45. Kosatsky T. The 2003 European heat waves. Eurosurveillance. 2005;10: 3–4.
- 46. Lowe D, Ebi KL, Forsberg B. Heatwave early warning systems and adaptation advice to reduce human health consequences of heatwaves. International journal of environmental research and public health. 2011;8: 4623–4648. pmid:22408593
- 47. Dong W, Zeng Q, Ma Y, Li G, Pan X. Impact of heat wave definitions on the added effect of heat waves on cardiovascular mortality in Beijing, China. International journal of environmental research and public health. 2016;13: 933. pmid:27657103
- 48.
Stocker TF, Qin D, Plattner G-K, Tignor MM, Allen SK, Boschung J, et al. Climate Change 2013: The physical science basis. contribution of working group I to the fifth assessment report of IPCC the intergovernmental panel on climate change. 2014.