Next Article in Journal
Coupling MATSim and the PALM Model System—Large Scale Traffic and Emission Modeling with High-Resolution Computational Fluid Dynamics Dispersion Modeling
Previous Article in Journal
Comprehensive Comparison of Seven Widely-Used Planetary Boundary Layer Parameterizations in Typhoon Mangkhut Intensification Simulation
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Atmosphere
Volume 15
Issue 10
10.3390/atmos15101184
atmosphere-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Microclimatic Variability and Thermal Comfort of Spectators in an Outdoor Stadium Venue

by
Andrew Collins
1,
Michael Brown
1,
Barrett Gutter
1 and
Christopher Fuhrmann
2,*
1
Department of Geosciences, Mississippi State University, Mississippi State, MS 39762, USA
2
NOAA’s Southeast Regional Climate Center, Department of Geography and Environment, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA
*
Author to whom correspondence should be addressed.
Atmosphere 2024, 15(10), 1184; https://doi.org/10.3390/atmos15101184
Submission received: 5 August 2024 / Revised: 25 September 2024 / Accepted: 28 September 2024 / Published: 30 September 2024
(This article belongs to the Section Biometeorology and Bioclimatology)
Browse Figures
Versions Notes

Abstract

:
This study examines heat exposure and its impact on the thermal comfort and health of spectators within a semi-outdoor American college football stadium in the southeastern United States. Over 50 sensors were deployed during the 2016 season from late August to late November to measure temperature and humidity across various stadium locations. Significant variations in temperature, heat index, and a modified version of the physiological equivalent temperature (mPET) were found within the stadium, with some areas exceeding National Weather Service heat alert thresholds during certain games. Moreover, mean temperatures in the stadium were higher than those measured at a nearby weather station, while the mean heat index was higher in the seating areas than in other stadium locations and at the nearby weather station. Reductions in modeled wind speed resulted in significant decreases in thermal comfort and greater physical stress among spectators, particularly when the wind was calm. Heat-related illness comprised up to two-thirds of all cases treated by first aid and emergency medical services during particularly hot games. Most of these occurred in the most thermally oppressive parts of the stadium. These results highlight the need for greater monitoring of heat exposure inside stadiums, earlier implementation of heat action plans to raise awareness and educate spectators on heat mitigation strategies, and incorporation of stadium design modifications that improve circulation, increase shade, and reduce crowding.

1. Introduction

Extreme heat is one of the leading causes of weather-related morbidity and mortality globally [1]. Previous research has identified several populations that are particularly vulnerable to extreme heat. These include outdoor workers, urban residents, the elderly, children, individuals with chronic diseases, individuals who are socially isolated, military personnel, and various racial and ethnic groups (see [2] for more details). Athletes are also vulnerable to extreme heat due to physical exertion and frequent outdoor exposure, which has led to the development of several heat mitigation strategies and safety guidelines for athletes [3].
Athletes associated with organized sports often perform in front of a wide range of spectators, sometimes in excess of 100,000 individuals. Since spectators may be exposed to the same environmental conditions as athletes and exert themselves through various behaviors (e.g., cheering), they are likewise vulnerable to the effects of extreme heat. Spectators are also vulnerable because they typically gather in dense crowds, consume food and beverages that contribute to dehydration, and generally possess lower levels of fitness and overall health compared to athletes, which may inhibit thermoregulation. In addition, spectators who travel, particularly long distances, to sporting events may not be acclimatized to the local conditions [4,5,6,7,8,9].
Since sporting events often exceed the typical threshold for a mass gathering (i.e., at least 1000 people), special planning is needed to address potential hazards, including extreme heat, and ensure access for emergency medical personnel [10]. Heat-related illness has been identified as a primary cause of morbidity and mortality at mass gatherings, including sporting events [11,12,13]. Previous studies have found that the apparent temperature (i.e., heat index) is a useful predictor of the number of people seeking medical attention for heat-related illnesses during these events (e.g., [14,15,16,17]).
In the United States, one of the most popular sporting events is American-style college football, which is played mostly in large outdoor and semi-outdoor stadiums during warm and humid times of the year (i.e., late summer and early fall). Previous studies of college football games found that increases in air temperature and heat index resulted in greater numbers of spectators seeking medical attention [14,15]. However, the meteorological observations from these studies were of conditions outside the stadium, which can be different from those inside the stadium, resulting in potential exposure misclassification (e.g., [18,19]). Indeed, studies of outdoor and semi-outdoor spaces, including stadiums, have noted higher temperatures [20,21,22] and greater absorption of radiation [23] compared to the outside environment. Other studies have placed meteorological sensors inside stadiums and compared the resulting temperatures to those from a nearby weather station. Gutter [24] and Reddick and Vanos [25] found that temperatures inside football stadiums at the University of Alabama and Texas Tech University, respectively, were several degrees warmer during individual games than temperatures recorded at a nearby automated weather station. These studies also found intra-stadium temperature variability related to small-scale changes in exposure to solar radiation, humidity, and the characteristics of the playing surface (i.e., natural grass vs. artificial turf).
Variations in temperature and the thermal comfort of spectators have been linked to aspects of stadium design. For example, stadium geometry and orientation can affect wind velocity and exposure to solar radiation [26]. While shading can reduce sun exposure, it can also decrease ventilation, resulting in thermal discomfort [27]. The configuration of seats and other gathering spaces within a stadium can lead to high crowd density, which can limit access to shade and reduce ventilation. It can also limit mobility and prevent access to hydration and cooling stations, as well as medical services [28]. Moreover, the energy gained from increased metabolic heat production can add to the heat load already induced by the ambient environment and enhanced by stadium materials (e.g., concrete, aluminum), design, and orientation [29]. These factors increase the likelihood that a spectator may experience heat-related illness.
In general, the effects of extreme heat on spectators at sporting events, such as American college football, are not well known, particularly in comparison to other heat-vulnerable populations. Moreover, there are very few in situ measurements from inside football stadiums that can determine where the most oppressive conditions are found, thus helping emergency managers better understand where the risk for heat stress is greatest so that resources to mitigate and respond to these risks can be properly allocated. Therefore, the objectives of this study are as follows: (1) describe the variability in temperature and humidity within a semi-outdoor American college football stadium; (2) measure the resulting thermal comfort of spectators during thermally oppressive games within the stadium; and (3) examine the relationship between measures of thermal comfort and reports of heat-related illness within the stadium during thermally oppressive games.

2. Materials and Methods

2.1. Data Collection

This study was conducted in Davis Wade Stadium on the campus of Mississippi State University in Starkville, MS, USA (33.450° N, 88.818° W). The stadium is used primarily for American-style football games and had a capacity of 61,337 at the time of this study [30]. The stadium is open-sky and characterized as “U-form” or horseshoe [26] with one open side, a continuous lower tier, and a discontinuous upper tier (Figure 1). The town of Starkville is located in the southeastern United States, which experiences a humid subtropical (Cfa) climate characterized by hot and humid summers and mild to cool winters with no distinct dry season.
To measure the temperature and humidity of the stadium, iButton sensors were attached to small plastic fobs and secured with plastic zip ties to railings throughout the stadium at heights ranging from 0.5 to 1.2 m above the ground (Figure 2). Two types of iButtons were used: Thermochrons (DS1921G-F5), which measure temperature only, and Hygrochrons (DS1923), which measure both temperature and relative humidity. Each iButton has a thermal response time of approximately 2 min, while the accuracy of the iButtons range from ±1.0 °C for Thermochrons to ±0.5 °C for Hygrochrons [31].
For this study, a total of 52 iButtons (40 Thermochrons and 12 Hygrochrons) were placed within the stadium (Figure 3). Each iButton was set to record continuously at 10 min intervals throughout the 2016 football season (29 August to 22 November). This recording interval was chosen to more precisely measure the duration of oppressive thermal conditions within the stadium during individual games (described below). Similar to [24,25], iButton observations were summarized across specific regions of the stadium.
  • Seating—spectator seating, which includes metal benches and plastic bleacher seats; most of these seats are not shaded; lower-tier seating is located on the east, west, and north sides of the stadium; upper-tier seating is located on the east and west sides of the stadium;
  • Field—ground-level; natural grass surface; east, west, and north sides are mostly unshaded; the south side is mostly shaded by a large video board (Figure 1);
  • Concourse—semi-outdoor environment that is shaded by stadium structures located on both the lower and upper tiers of the stadium.
As in [24,25], comparisons were made between seating, field, and concourse temperatures and those from a nearby automated weather station. The station closest to Davis Wade Stadium is a USDA-NRCS Soil Climate Analysis Network (SCAN) station (#2064) located approximately 2 km to the northeast in a university-owned agricultural field (https://wcc.sc.egov.usda.gov/nwcc/site?sitenum=2064&temp_unit=9, [32] accessed on 12 June 2020). Because the SCAN station only records hourly temperatures, all 10 min iButton observations were converted to hourly values by calculating the average of all observations from the previous hour. For example, all 10 min observations occurring between 3:00 p.m. and 3:59 p.m. were averaged to produce a single 3:00 p.m. hourly observation. This hourly observation was then compared to the corresponding hourly observation from the SCAN station.
It is important to note that all iButtons placed in the seating and field level areas were not shielded, though they were oriented face-down to limit exposure to direct sunlight (as shown in Figure 2). A study by Bailey et al. [33] found that, while highly correlated (r = 0.96), daytime temperatures recorded by unshielded iButtons were warmer than temperatures recorded by the SCAN station over a two-week period in September 2018, which is seasonally consistent with the present study. Another study conducted during the summer in Baltimore, MD, USA, also found that daytime temperatures recorded by unshielded iButtons were several degrees warmer than those recorded by a nearby weather station [34]. Therefore, to facilitate comparisons between iButton and SCAN temperatures, hourly average temperature differences between downward facing iButtons and the SCAN station reported in Bailey et al. [33] were applied as correction factors to all iButton observations in the seating and field level areas of the stadium (Figure S1). Only those differences exceeding the accuracy of the iButton sensors (±1.0 °C) were applied as correction factors. These occurred between 7:00 a.m. and 4:00 p.m., when iButtons were warmer than the SCAN station (average difference of 3.5 °C), and between 7:00 p.m. and 9:00 p.m., when iButtons were cooler than the SCAN station (average difference of 1.4 °C). As an example, if an iButton recorded a temperature of 42 °C at 3:00 p.m. (15 LST), a correction factor of 3 °C was applied (as shown in Figure S1), resulting in a corrected temperature of 39 °C. Measures of agreement reported in Bailey et al. [33], including mean absolute error (MAE = 2.11), mean bias error (MBE = 0.44), root mean square error (RMSE = 3.30), and Lin’s concordance correlation coefficient (Pc = 0.90), reveal that iButtons more accurately and precisely approximate SCAN temperatures than other sensors (i.e., HOBO pendant, Kestrel DROP), confirming the reliability of our approach. Additionally, while unshielded iButtons will naturally have a temperature bias over shielded instruments, the effect is minimized with the Hygrochron models due to a filter on the lid that promotes airflow over the sensor. This not only allows water vapor to pass over the hygrometer sensor (to record humidity) but helps remove heat from the sensors in the same way that an aspirator removes heat from a shielded thermometer [31].
In addition, to ensure that all devices were operating properly, temperatures measured within 30 min prior to sunrise were compared across all iButtons in the stadium (seating, field, and concourse) during the study period. The period just prior to sunrise eliminates the influence of solar radiation, resulting in more homogeneity in temperature observations [35]. Variations in mean temperatures across all iButtons were within the accuracy of the sensors (±1.0 °C), indicating no malfunctioning devices or erroneous measurements.

2.2. Calculation of Thermal Indices

To assess the thermal stress and comfort of stadium spectators, two indices were calculated using temperature and humidity observations from the Hygrochron iButtons: Heat Index and Physiological Equivalent Temperature. The Heat Index (HI) is a screening tool that provides a quick but less accurate assessment of how hot it feels to the human body by accounting only for air temperature and humidity [36,37]. It is used in the United States to issue heat alerts and provide guidance on heat-health warning systems. The Physiological Equivalent Temperature (PET) is a more expert measure of thermal comfort that combines environmental and physiological factors into a single index. Unlike the HI, it also considers wind speed, heat transfer by radiation (solar and infrared emitted from surfaces), the metabolic rate, and clothing insulation. Additionally, PET accounts for the body’s heat exchange processes, including radiation, convection, and evaporation, which reflect how humans experience the surrounding microclimate.
The HI was calculated from temperature and relative humidity observations from the Hygrochron iButtons and SCAN station using the equation adopted by NOAA’s National Weather Service (NWS):
H I = − 42.379 + 2.04901523 T + 10.14333127 R − 0.22475541 T R   − 6.8378 × 10 − 3 T 2 − 5.481717 × 10 − 2 R 2   + 1.22874 × 10 − 3 T 2 R + 8.5282 × 10 − 4 T R 2   − 1.99 × 10 − 6 T 2 R 2
where T is the temperature (°F) and RH is the relative humidity (percentage). Since this equation is not suitable for temperatures below 80 °F (26.7 °C), the following Rothfusz [38] equation was used to recalculate all HI values below 80 °F:
H I = 0.5 × { T + 61.0 + T − 68.0 × 1.2 + R H × 0.094 }
Both 10 min and hourly averaged HI values were used to assess the thermal stress of spectators across the different sections of the stadium (i.e., seating, field level, and concourse), while only hourly HI values were calculated from the SCAN station and used to compare with the corresponding hourly iButton values.
In this study, RayMan Pro was used to calculate the PET of stadium spectators [4,39,40]. PET is based on the global thermal balance of the human body, which is calculated using the Munich energy model for individuals, or MEMI [41]:
M + W + C + R + E D + E R E + E S W = 0
where M is metabolism, W is energy losses, C is convective heat transfer to the environment, R is radiative heat transfer to the environment, ED is perspiration, ERE is respiration, and ESW is heat flow due to evaporation of sweat with all variables expressed in Watts [42]. In computing the conductive and convective heat transfer terms, it is assumed that the spectators are standing, which is common practice during American collegiate football games.
The calculation of PET involves both environmental and physiological parameters. One of the environmental parameters is the mean radiant temperature (Tmrt), which accounts for all radiant fluxes (shortwave and longwave) experienced by a human [43]:
T m r t = ( S s t r / ( ε p σ 4 ) ) − 273.15
where Sstr is the mean radiant flux density, εp is the emissivity of the human body (standard value of 0.97), and σ is the Stefan-Boltzmann constant (5.67 × 10−8 Wm−2 K−4). In urban areas, tall buildings and other structures can modify the radiant fluxes and create uneven patterns of heating, referred to as urban canyons. Matazarakis et al. [43] found that PET was significantly lower along highly shaded streets characterized by a deep urban canyon. Like urban areas, open-sky or semi-outdoor stadiums often exhibit complex geometry that can resemble an urban canyon (see Figure 1). The depth of an urban canyon, or more generally, the amount of obscuration from buildings and other structures, can be determined by calculating the sky view factor (SVF). The SVF is a dimensionless value from 0 to 1 where 0 is complete obscuration and 1 is no obscuration (i.e., clear sky). The SVF can be included in the calculation of PET in RayMan Pro. In this study, circumpolar fisheye photographs were taken next to the Hygrochron iButtons (Figure 3B) to determine the fraction of visible sky from four horizontal directions (north, south, east, and west). Due to an ingest problem, the images were unable to be imported directly into RayMan Pro. Instead, the fraction of visible sky in each image was calculated through digitization, and the values were manually entered into RayMan Pro. The SVF values at field level ranged from 0.78 in the south end zone to 1.0 in the north end zone and at the center of the field (average of 0.88). Values in the seating areas ranged from 0.78 above the north end zone to 0.98 in the lower east corner (average of 0.89). An SVF value of 0 (complete obscuration) was used to calculate PET in the concourses.
Additional environmental parameters used to calculate PET in RayMan Pro include air temperature, relative humidity, cloud cover, and wind speed. Temperature and relative humidity were obtained from Hygrochron iButton observations within the stadium (Figure 3B). Cloud cover was set at 2 octas (i.e., few clouds, or 25% cloud cover), which is consistent with observations made during the games examined. Although the SCAN station records hourly wind speed, it is not known whether these observations are representative of wind speeds within the stadium. Therefore, various wind speed thresholds were tested using RayMan Pro. The specific thresholds, based on modeling results from [44], were 0.1 ms−1, 2.1 ms−1, 4.1 ms−1, and 6.1 ms−1. In this study, we used a modified version of PET (mPET) that more accurately accounts for the effects of humidity and variations in clothing [4,45,46]. Specifically, mPET incorporates a more complex series of body and clothing models that better represent sensible and latent heat transfers inside and outside the body. These improvements have been shown to provide more conservative measures of thermal comfort than PET in humid subtropical climates (such as the one in this study), which, if adopted for heat mitigation, would offer guidance that errs on the side of caution [4]. Regarding the physiological inputs used to calculate mPET, the metabolic activity level of a spectator that is “very excited, emotional, and cheering” was set at 166 Wm−2 [19,47]. Other physiological inputs included gender (male), age (35 years old), weight (75 kg), height (1.5 m), and clothing insulation (clo = 0.33) [19,20,41,48,49,50,51].

2.3. First Aid and EMS Data

First aid and emergency medical services (EMS) data for select games were obtained from the Office of Emergency Information at Mississippi State University. Parameters included the time of medical treatment, the general location of the patient within the stadium (e.g., lower-tier seating on the east side), the chief complaint (e.g., heat-related), and final disposition (e.g., transferred to hospital, further care refused). The distribution of heat-related cases within the stadium was compared to their corresponding (10 min) HI and mPET values in an effort to relate thermal stress and comfort to the location of observed heat-related health outcomes. In this way, the use of HI and/or mPET values observed within the stadium for emergency management and stadium operations could be evaluated alongside actual cases of heat-related illness.

2.4. Statistical Methods

To compare hourly temperature, HI, and mPET values across the different stadium locations and SCAN stations, χ2 tests were performed to determine if the values were normally distributed. Since the distributions failed the tests for normality, bootstrapping was used. Bootstrapping is a non-parametric technique that does not require datasets to be normally distributed or have the same distribution. Bootstrapping uses resampling, with replacement, to generate quantile values for statistical measurements, such as the mean and standard deviation. A total of 1000 bootstrap replicates were created from hourly temperature, HI, and mPET values across the different stadium locations and the SCAN station. These bootstrap replicates were tested at the 95% confidence interval to determine if the mean and standard deviation of the different stadium observations fell within the distribution of bootstrap replicates from the SCAN station and the other stadium locations. The null hypothesis in this case was that the means and standard deviations were equal, while the alternative hypothesis was that the means and standard deviations were not equal. The null hypothesis was rejected in cases where the mean or standard deviation of one of the distributions (e.g., seating) did not fall within the first and third quantiles of one of the other distributions (e.g., SCAN station).

3. Results

3.1. Thermal Profile of the Stadium

A summary of mean temperature and HI observations within the stadium and at the SCAN station is presented in Table 1. Temperatures across each region of the stadium were measured using both the Thermochrons and Hygrochrons, while HI was measured using the Hygrochrons only. During the study period, mean temperatures within each of the three areas of the stadium were statistically higher than the mean temperature recorded at the SCAN station. Within the stadium, the mean temperature in the concourses was statistically higher than the mean temperatures in the seating and field areas. However, the mean concourse temperature was statistically less variable than those in the other areas of the stadium and at the SCAN station. The mean HI during the study period was statistically higher and more variable in the seating area of the stadium compared to the field area and SCAN station. In contrast, the mean HI value in the concourses was statistically lower and less variable than the rest of the stadium and the SCAN station. Despite a few statistically significant differences, the results in Table 1 reveal that mean temperatures and HI values averaged across the four-month study period vary minimally (i.e., within 2 °C) across locations. A more complete picture of temperature variability within the stadium can be obtained by mapping observations from individual iButtons at different times of the day. The resulting thermal maps of the stadium allow for the identification of areas where exposure to extreme heat may be elevated.
Plots of stadium temperatures by hour for two games during the study period are shown in Figure 4 (29 October 2016) and Figure 5 (10 September 2016). Each figure displays the temporal evolution of temperature for the duration of each game, as well as the preceding and succeeding hours when most spectators were entering and exiting the stadium, respectively. The game on 29 October was played during the afternoon when temperatures are typically highest (2:00 p.m. to 5:00 p.m. LST), while the game on 10 September was played during the early evening following the peak in daily temperature (6:00 p.m. to 9:00 p.m. LST).
The diurnal variability in temperature within the seating and field areas of the stadium is largely a function of solar position, which is defined by the elevation of the sun above the horizon (i.e., solar elevation angle) and the compass direction (degrees clockwise from north) from which the sun is shining (i.e., azimuth angle). One hour prior to the game on 29 October, the sun reached its highest local elevation of the day (43°) with an azimuth of 167°, meaning the sun’s most intense and direct rays were shining onto the west side of the stadium (Figure 4A). As a result, temperatures on this side of the stadium were as much as 3 °C higher than temperatures on the opposite (i.e., east) side of the stadium. Compared to the SCAN station, the average temperature in the seating and field areas was over 4 °C higher, with the greatest differences (10 °C) found in the northwest corner of the seating area. Average and maximum temperature differences were slightly higher by game time (5 °C and 11 °C, respectively), and while the west side of the stadium was still warmer than the north and east sides, the differences between them were smaller (1–2 °C) (Figure 4B).
One hour after the game began, the SCAN station recorded a daily maximum temperature of 30.6 °C. The average temperature in the seating and field areas was again about 4–5 °C higher than the SCAN station, but there was much greater variability, particularly on the west side (almost 9 °C among iButton observations) (Figure 4C). The highest temperature (42.5 °C) was recorded in the seating area on the north side of the stadium, which was receiving the most direct and intense solar radiation (solar elevation of 39° and azimuth of 187°). It is noteworthy, however, that temperatures exceeding 37.8 °C were recorded by multiple iButtons in all seating areas of the stadium. The effect of the large video board on the south field level of the stadium is also apparent, as temperatures recorded in its shadow were almost 7 °C cooler than temperatures recorded throughout much of the seating areas.
Over the next two hours, the azimuth angle increased from 206° to 244°, meaning the sun’s most intense and direct rays were shining onto the east side of the stadium. As a result, average (maximum) temperatures on this side of the stadium were 3–4 °C (10 °C) higher than temperatures on the west side and at the SCAN station (Figure 4D,E). Despite a notable drop in the solar elevation angle over this time (32° to 12°), several iButtons on the east side of the stadium recorded temperatures between 32–38 °C. One hour after the game (Figure 4F), the solar elevation was very close to the horizon (1°), and temperature variability within the seating and field areas was about 1–2 °C lower than that observed over the previous hours. However, temperatures in these areas cooled much more slowly than those at the SCAN station, resulting in an average temperature difference of nearly 6 °C.
Intra-stadium temperature variability, as well as marked differences in stadium and SCAN station temperatures, were also observed during the game on 10 September, particularly at the start of the game and during the hour prior (Figure 5A,B). The highest temperatures (39–42 °C) were recorded in the seating area on the north and east sides of the stadium, which were receiving the most direct and intense solar radiation (solar elevation of 38–26° and azimuth of 257–266°). These temperatures were over 6 °C higher than the temperatures recorded at the SCAN station and on the west side of the stadium. As was seen in the game on 29 October, intra-stadium temperature variability decreased as the solar elevation angle approached zero (Figure 5C,D), with seating and field level temperatures varying by only 1–2 °C after astronomical twilight (Figure 5E,F). Moreover, stadium temperatures cooled more slowly than those recorded at the SCAN station, resulting in temperature differences of 2–5 °C up to an hour after the game.
Observations from these two games revealed notable differences in temperatures between the stadium and the surrounding environment, as well as smaller-scale differences within the stadium. When averaged over the full study period, iButton temperatures on the north end of the stadium (field level and seating) were statistically higher than temperatures recorded on all other sides of the stadium. Due to the large video board casting a shadow onto the south end of the field, iButton temperatures there were 3–5 °C lower on average than the rest of the stadium. Additionally, hourly variations in temperature were greater at field level than in the seating areas.

3.2. Thermal Comfort of Stadium Spectators

Two metrics were calculated to assess the thermal stress and comfort of stadium spectators: HI and mPET. Because these metrics incorporate humidity, only the Hygrochron iButtons were used. Conditions in the stadium and at the SCAN station were related to the likelihood of heat stress using established ranges and associated alert categories for both metrics (Table 2).
Figure 6 shows the hourly averaged HI values across the different sections of the stadium (see locations in Figure 3B) and the corresponding values at the SCAN station for three games during the study period: 3 September 2016 (Figure 6A), 10 September 2016 (Figure 6B), and 29 October 2016 (Figure 6C). These games were chosen out of the six that took place over the study period because over half of the hourly averaged HI observations in the stadium (out of 72 total observations) were at least 26.7 °C, which is the threshold for the lowest heat stress category based on HI (Table 2), and because stadium capacity was at least 95%, resulting in crowding that could exacerbate heat stress. As such, these games were considered thermally oppressive and heat-stressful for most spectators and are examined in more detail in the following two sections.
A total of 60 HI observations were made in the seating and field areas during two of the three games, while 54 HI observations were made during the third game due to the theft of the Hygrochron in the upper-tier seating area on the west side of the stadium. The most thermally oppressive game was played on 3 September. More than half (53%) of the HI observations in the seating and field areas reached at least the Danger category, where heat exhaustion is likely, and heat stroke is possible with prolonged exposure (Figure 6A; Table 2). Two observations in the lower-level seating area on the northeast side of the stadium reached the Extreme Danger category, with HI values of 53.9 °C and 59.4 °C. These were the highest HI values recorded in the stadium during the study period and represent conditions where heat stroke is likely with prolonged exposure. In contrast, HI values recorded by the SCAN station were mostly in the Extreme Caution category, while HI values in the concourses were in the Caution category (Figure 6A). The most oppressive conditions in the stadium were recorded at halftime at 1:00 p.m. LST. The average HI in the seating (46.1 °C) and field (40.6 °C) areas was much higher than the average HI in the concourses (30.6 °C) and the HI recorded at the SCAN station (33.9 °C).
Conditions in the stadium during the other two games also reached the Danger category, but for a shorter period of time and at far fewer locations. During the game on 10 September, HI values in the seating area on the east side of the stadium ranged from 40.6 °C to over 43.3 °C, while the rest of the seating and field areas were in the Extreme Caution category. HI values in the concourses and at the SCAN station were also in the Extreme Caution category during the first half of the game (from 5:00 p.m. to 7:00 p.m. LST). Due to a nearby thunderstorm, HI values dropped below the Caution category (26.7 °C) at the SCAN station and across much of the stadium during the latter part of the game. In general, the differences in HI values between the stadium and the SCAN station were smaller during this game than the other two games.
The game played on 29 October may be considered the most thermally oppressive given the time of year, particularly for spectators who were not acclimatized earlier in the season. Remarkably, HI values in the seating area on the east side of the stadium ranged from 40 °C to 45.6 °C (Danger category), while several other locations were in the Extreme Caution category. In contrast, HI values recorded at the SCAN station were in the Caution category (28–29 °C). The most oppressive conditions in the stadium were recorded at the start of the game at 2:00 p.m. LST. As was the case during the first game, the average HI in the seating (37.2 °C) and field (35 °C) areas were much higher than the average HI in the concourses (26.7 °C) and the HI recorded at the SCAN station (33.9 °C).
Table 3 provides a summary of mean mPET values within the stadium and at the SCAN station for each of the four wind scenarios. These values are consistent with those calculated in similar climates [4,53]. During the study period, the mean mPET was higher and more variable in the seating area compared to the rest of the stadium and the SCAN station. These differences were statistically significant at the lowest wind threshold, where conditions were classified as Hot and the grade of physical stress classified as Strong. In contrast, the mean mPET was statistically lower and less variable in the concourses than the rest of the stadium and the SCAN station across all four wind thresholds. Conditions in the concourses were classified as Slightly Warm with only a Slight grade of physical stress. When examined across the wind thresholds, mean mPET was statistically higher and more variable at the lowest threshold compared to the higher thresholds across all parts of the stadium and at the SCAN station. Differences in mPET were smaller and less variable across the three higher wind thresholds, suggesting that just a slight increase in wind can have a significant effect on thermal comfort. This is further revealed when examining mPET values during individual games. When using the lowest wind threshold, Very Hot conditions and an Extreme grade of physical stress were observed in the stadium during the entire duration of the game on 3 September. However, the duration of these conditions decreased to just one hour when the wind threshold was increased to 4.1 ms−1. Similarly, during the game on 29 October, an increase in wind from 0.1 ms−1 to 4.1 ms−1 reduced mPET in the stadium from Hot conditions and Strong heat stress for the majority of the game to Slightly Warm and Moderate heat stress for more than half of the game.

3.3. Thermal Comfort and Heat-Related Illness

First aid and emergency medical services (EMS) data were analyzed alongside corresponding HI and mPET values for the three games discussed in the previous two sections (Table 4). As in prior studies [14,15], the number of individuals seen by first aid and EMS was standardized by the reported attendance at each game. The number of individuals seen for medical care was highest during the game on 3 September (8.4 per 10,000 spectators). This number decreased by half during the next game on 10 September (4.2 per 10,000 spectators). The number of individuals seen for medical care was lowest during the game on 29 October (1.7 per 10,000 spectators). The percentage of cases where the chief complaint was listed as heat-related was 42% during the game on 3 September, 63% during the game on 10 September, and 60% during the game on 29 October. The majority of these cases were noted on the east side of the stadium during all three games, which was more thermally oppressive than the west side of the stadium according to the mean HI and mPET in the seating areas.

4. Discussion

This study used over 50 iButton sensors to measure the temperature and humidity within a semi-outdoor American college football stadium in the southeastern United States during an approximately four-month period coinciding with the football season of 2016. These observations were also used to calculate thermal indices related to heat stress and thermal comfort of spectators in the stadium. The results were corroborated by first aid and EMS data on heat-related illnesses. While previous studies have related health outcomes in stadiums to temperature or HI, few have used observations from within the stadium. Instead, observations from a nearby weather station are used to represent conditions inside the stadium [14,15], which can result in exposure misclassification. Also, few studies have modeled the thermal comfort of spectators, a vulnerable population that has been largely understudied in the heat literature.
Those in charge of game day operations rely on forecasts of meteorological conditions to assess potential hazards. However, these forecasts are for conditions outside the stadium, which, as shown in this study and others [20,22,24,25], can be very different from conditions inside the stadium. In this study, several locations in the stadium were found to be thermally oppressive, placing spectators and other stadium occupants at risk for heat-related illness. Moreover, heat exposure can be very localized, with temperatures varying by as much as 10 °C within the seating areas of the stadium. Relying on observations from the nearest weather station or from official weather forecasts would not have revealed this hazard. Instead, by mapping conditions across various parts of the stadium, game day officials and emergency managers can better identify where exposure to heat is the greatest and allocate additional staff and resources to those areas. These maps, along with real-time observations, can also be used by local meteorologists to communicate the heat hazard within the stadium (Figure 7), as well as inform design modifications and renovations to reduce heat exposure.
HI values were highest in the seating areas of the stadium. Of the three games studied in detail, which were considered thermally oppressive, multiple locations reached the Danger category, including the game on 29 October, while a few locations reached the Extreme Danger category during the game on 3 September. Interestingly, conditions in the concourses during this game only reached the Caution category, which were actually less oppressive than those outside the stadium (as measured by the SCAN station). While the concourses observed the highest mean temperature in the stadium during the study period, they were the least oppressive locations according to both thermal indices (HI and mPET). This is likely due to the lack of direct solar radiation, lower humidity, and adequate circulation, as well as less crowding of spectators compared to the seating areas. Crowding of spectators can increase heat and humidity, particularly through perspiration, and reduce wind speed, thereby reducing evaporative and convective heat losses from the body [5,29]. As the reported attendance for each of the three games was around 95%, with most of the unoccupied seats located in the upper tiers of the stadium, crowding of spectators may at least partially explain the exceptionally high HI values observed in some of the lower seating areas (>50 °C).
The NWS uses forecasts and observations of HI to issue heat alerts, namely Heat Advisories and Excessive Heat Warnings [54]. In many instances, these alerts, which are based on locally-specific thresholds, will trigger public health and emergency response plans that target vulnerable populations. It is notable that conditions measured at the SCAN station did not reach the Advisory level (40.6 °C) during any of the three thermally oppressive games. However, during the game on 3 September, nearly half of the locations measured in the stadium reached the Advisory level, while about a quarter of these locations reached the Warning level (43.3 °C). Remarkably, during the game on 29 October, three locations on the east side of the stadium reached the Warning level, with HI values over 15 °C higher than those recorded outside the stadium. A recent study of HI criteria used by the NWS found that current values associated with the Danger and Extreme Danger categories, which approximately coincide with Advisory and Warning thresholds, may, in fact, be too low due to incorrect modeling of thermoregulation at such high temperature and humidity combinations [55]. This underscores the need for proper communication, monitoring, safety guidelines, heat action protocols, and design modifications to safeguard spectators from heat-related illnesses.
In addition to HI, it is useful to consider thermal indices that account for other meteorological variables, as well as physiologic variables that are related to human energy balance. In this study, mPET was modeled at the same stadium locations as HI and at the SCAN station. mPET is a thermal comfort index that accounts for spectator exposure, the effects of crowding, and physical exertion. Because evaporation and convection are the primary mechanisms by which the body cools, mPET was modeled using different wind speeds. Mean mPET values revealed that most locations in the stadium present some level of thermal discomfort, even at higher modeled wind speeds. Conditions were classified as Hot in the seating and field areas and Slightly Warm in the concourses. However, during individual games, Very Hot conditions were noted with an Extreme grade of physical stress at lower modeled wind speeds. Similar levels of physical stress were noted among spectators in the soccer stadium during matches played as part of the 2015 Pan American games in Toronto, Canada [19]. In general, mPET values were higher at lower wind speeds, which is consistent with other studies [20,26,41,48,50]. However, mPET was also more variable at lower speeds, suggesting that other factors may be contributing to mPET when the wind is calm. Nevertheless, any wind, even a small breeze, can significantly improve thermal comfort under most conditions.
First aid and EMS data revealed that between about 40–60% of cases were for heat-related illness, with most occurring on the most thermally oppressive side of the stadium. The association between higher temperatures, heat indices, and an increased number of patients treated by first aid and EMS has been noted in other studies of American college football stadiums [14,15,25]. However, these studies did not specifically examine patients treated for heat-related illness. The greatest number of patients treated for heat-related illness was found during the game on 3 September, which was the first game of the season and also the warmest and most thermally oppressive. The game on 29 October saw the fewest number of patients treated for heat-related illness despite unseasonably warm temperatures and HI values that exceeded heat alert thresholds in several parts of the stadium. It is likely that spectators had greater awareness of the heat hazard by this time in the season and, therefore, took preventative measures or perhaps were already acclimatized. This suggests that athletic departments should begin communicating heat action plans prior to the start of the season to increase awareness and educate spectators on heat mitigation strategies, including acclimatization.
Improving thermal comfort within a stadium can also be accomplished through various design modifications. In humid climates like the southeastern U.S., one of the most effective ways to improve thermal comfort is to increase airflow, as demonstrated by modeling of mPET using different wind scenarios. Other considerations include reducing exposure to solar radiation (which is tied to solar position relative to stadium orientation), reducing humidity, and reducing the amount of heat emitted from stadium materials. It is important to consider that design modifications that diminish one factor may amplify another. For example, while roofs, awnings, and shade sails have been shown to reduce exposure to solar radiation [26], they can also impede airflow. An examination of the design of the stadium used for the 2000 Olympics in Sydney, Australia, found that potential heat risk was actually highest in the shaded part of the stadium due to reduced airflow and ventilation, as well as increased heat load due to the use of semi-transparent roofing material [27]. Additionally, while using natural instead of artificial turf (as is the case in this study) can help lower temperatures at the field level, it can increase the amount of humidity. The same effect may occur due to increased use of vegetation and green space in other parts of the stadium [56].
Universities routinely update and renovate their athletic facilities, which sometimes involves constructing an entirely new facility. These projects should consider the thermal comfort of all stadium occupants and how it can be maximized with various design features. With respect to spectators, athletic departments should prioritize access to open areas to improve airflow and reduce crowding, increase the number of climate-controlled spaces within the stadium, and expand access to shade. It is important to keep in mind, however, that some design modifications may improve thermal comfort for some occupants but decrease it for others. In this case, improvements in thermal comfort for spectators should not lead to a decrease in thermal comfort for players, coaches, referees, and other individuals on the field or working in the stadium. Ultimately, stadiums that are completely enclosed and climate-controlled offer the best opportunity to create thermally optimal conditions for all occupants, but such a design, especially for a university stadium, is often not economically viable.
Another way to maximize thermal comfort and limit heat stress in stadiums is to schedule games during less oppressive times of the day, such as in the evening. However, moving games to the evening may result in a delay in heat-related illnesses due to spectators tailgating in the hours prior to the game, which can increase their cumulative heat load due to prolonged exposure and consumption of food and beverages that lead to dehydration. Future work on heat safety in stadiums should include surveying spectator behaviors and meteorological conditions prior to the game, which may provide more insight into the factors that lead to heat stress. Additionally, while it is not feasible to reschedule an entire American college football season, some sporting events may consider scheduling during more thermally optimal times of the year. A notable example is the 2022 FIFA World Cup matches in Qatar, which were moved from the exceptionally hot months of June and July to the more thermally comfortable months of November and December [5].
It is important to note some limitations of this study. First, the results may only be generalizable to other average-sized stadiums (in terms of capacity and acreage) located in a humid subtropical climate. Similar studies conducted in stadiums of different sizes, designs, materials, field types, and orientations, as well as in different climates and physical environments, will likely yield different results. Second, ambient temperatures during the study period were over 3 °C warmer than the 30-year climatological mean temperature at the Cooperative Observer station in Starkville, which is located adjacent to the SCAN station. As such, the results of this study may represent an extreme set of scenarios of thermal comfort and heat stress in the stadium, while a season with more climatologically normal or below-average temperatures may result in fewer games and stadium locations with thermally oppressive conditions. Third, while first aid and EMS stations are a primary access point for individuals seeking medical attention, there are an unknown number of spectators who are unable or simply unwilling to access medical care. As a result, the number of patients reported in this study is likely an underestimation of the overall health burden associated with heat in the stadium. Fourth, the use of unshielded iButtons introduced some bias in the analysis, which we attempted to resolve by applying a correction factor to the hourly data. In addition, the iButton accuracy is inconsistent with ISO standards for thermal comfort measurements, which also introduced some bias. Finally, several assumptions were made in the modeling of mPET that, while based on prior studies and observed meteorological conditions, could be adjusted to represent variations in metabolic activity, body composition, age, as well as variations in cloud cover and wind speeds within the stadium. Future work could also examine additional thermal indices that better represent the effects of radiation than HI (e.g., wet bulb globe temperature, universal thermal climate index), as well as conduct surveys of spectator thermal perception to better understand the connection between thermal comfort, heat stress, and the microclimate of the stadium.

5. Conclusions

We found significant variations in heat exposure within a semi-outdoor American college football stadium, with some areas exceeding NWS heat alert thresholds. Mean temperatures inside the stadium were higher than those recorded at a nearby weather station, and the mean heat index was particularly elevated in the seating areas compared to other parts of the stadium and the weather station. When modeled wind speeds were reduced, thermal comfort (as measured by mPET) decreased significantly, leading to greater physical stress among spectators, especially when the wind was calm. During particularly hot games, up to two-thirds of all cases treated by first aid and emergency medical services were heat-related, with most occurring in the hottest areas of the stadium. These findings underscore the need for enhanced monitoring of heat exposure within stadiums, earlier activation of heat action plans to inform and educate spectators on heat mitigation strategies and resources (e.g., cooling stations, hydration points), and stadium design modifications to improve airflow, increase shade, and reduce crowding.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/atmos15101184/s1, Figure S1: Difference between mean iButton temperatures in the stadium and temperatures recorded at the SCAN station for each hour of the day (in local standard time) over the study period. Positive (negative) values indicate that iButton temperatures were higher (lower) than those at the SCAN station.

Author Contributions

Conceptualization, M.B., B.G. and C.F.; methodology, M.B., B.G. and C.F.; formal analysis, A.C., M.B., B.G. and C.F.; resources, M.B., B.G. and C.F.; data curation, A.C.; writing—original draft preparation, A.C. and C.F.; writing—review and editing, A.C., M.B., B.G. and C.F.; visualization, A.C. and B.G.; supervision, C.F.; project administration, C.F. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy.

Acknowledgments

Megan Williams, Mandy Raborn, and Crystal Worley for assisting with data collection; Brent Frey for access to the stadium; Brent Crocker for providing the first aid and EMS data; Jennifer Vanos for feedback on the design of this study and review of preliminary results.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Campbell, S.; Remenyi, T.A.; Whiteb, C.J.; Johnston, F.H. Heatwave and health impact research: A global review. Health Place 2018, 53, 210–218. [Google Scholar] [CrossRef] [PubMed]
  2. Kravchenko, J.; Abernethy, A.P.; Fawzy, M.; Lyerly, H.K. Minimization of heatwave morbidity and mortality. Am. J. Prev. Med. 2013, 44, 274–282. [Google Scholar] [CrossRef] [PubMed]
  3. Casa, D.J.; DeMartini, J.K.; Bergeron, M.F.; Csillan, D.; Eichner, E.R.; Lopez, R.M.; Ferrara, M.S.; Miller, K.C.; O’Connor, F.; Sawka, M.N. National Athletic Trainers’ Association position statement: Exertional heat illnesses. J. Athl. Train. 2015, 50, 986–1000. [Google Scholar] [CrossRef] [PubMed]
  4. Matzarakis, A.; Frohlich, D.; Bermon, S.; Adami, P.E. Quantifying thermal stress for sport events—The case of the Olympic Games 2020 in Tokyo. Atmosphere 2018, 9, 479. [Google Scholar] [CrossRef]
  5. Vanos, J.K.; Kosaka, E.; Iida, A.; Yokohari, M.; Middel, A.; Scott-Fleming, I.; Brown, R.D. Planning for spectator thermal comfort and health in the face of extreme heat: The Tokyo 2020 Olympic marathons. Sci. Total Environ. 2019, 657, 904–917. [Google Scholar] [CrossRef]
  6. Schuster, C.; Honold, J.; Lauf, S.; Lakes, T. Urban heat stress: Novel survey suggests health and fitness as future avenue for research and adaptation strategies. Environ. Res. Lett. 2017, 12, 44021. [Google Scholar] [CrossRef]
  7. De Freitas, C.R.; Scott, D.; McBoyle, G. A second generation climate index for tourism (CIT): Specification and verification. Int. J. Biometeorol. 2008, 52, 399–407. [Google Scholar] [CrossRef]
  8. Meehan, P.; Toomey, K.E.; Drinnon, J.; Cunningham, S.; Anderson, N.; Baker, E. Public health response for the 1996 Olympic Games. JAMA 1998, 279, 1469–1473. [Google Scholar] [CrossRef]
  9. Verdaguer-Codina, J.; Martin, D.E.; Pujol-Amat, P.; Ruiz, A.; Prat, J.A. Climatic heat stress studies at the Barcelona Olympic Games, 1992. Sport Med. Train. Rehabil. 1995, 6, 167–192. [Google Scholar] [CrossRef]
  10. Soomaroo, L.; Murray, V. Disasters at Mass Gatherings: Lessons from History. PLoS Curr. 2012, 4, RRN1301. [Google Scholar] [CrossRef]
  11. Arbon, P.; Bridgewater, F.H.; Smith, C. Mass gathering medicine: A predictive model for patient presentation and transport rates. Prehospital Disaster Med. 2001, 16, 150–158. [Google Scholar] [CrossRef] [PubMed]
  12. Milsten, A.M.; Maguire, B.J.; Bissell, R.A.; Seaman, K.G. Mass-gathering medical care: A review of the literature. Prehospital Disaster Med. 2002, 17, 151–162. [Google Scholar] [CrossRef] [PubMed]
  13. Shelton, S.; Haire, S.; Gerard, B. Medical care for mass gatherings at collegiate football games. South. Med. J. 1997, 90, 1081–1083. [Google Scholar] [CrossRef] [PubMed]
  14. Perron, A.D.; Brady, W.J.; Custalow, C.B.; Johnson, D.M. Association of heat index and patient volume at a mass gathering event. Prehosp. Emerg. Care 2005, 9, 49–52. [Google Scholar] [CrossRef] [PubMed]
  15. Kman, N.E.; Russell, G.B.; Bozeman, W.P.; Ehrman, K.; Winslow, J. Derivation of a Formula to Predict Patient Volume Based on Temperature at College Football Games. Prehosp. Emerg. Care 2007, 11, 453–457. [Google Scholar] [CrossRef] [PubMed]
  16. Baird, M.B.; O’Connor, R.E.; Williamson, A.L.; Sojka, B.; Alibertis, K.; Brady, W.J. The impact of warm weather on mass event medical need: A review of the literature. Am. J. Emerg. Med. 2010, 28, 224–229. [Google Scholar] [CrossRef]
  17. Steffen, R.; Bouchama, A.; Johansson, A.; Dvorak, J.; Isla, N.; Smallwood, C.; Memish, Z.A. Non communicable health risks during mass gatherings. Lancet Infect. Dis. 2012, 12, 142–149. [Google Scholar] [CrossRef]
  18. Bernhard, M.C.; Kent, S.T.; Sloan, M.E.; Evans, M.B.; McClure, L.A.; Gohlke, J.M. Measuring personal heat exposure in an urban and rural environment. Environ. Res. 2015, 137, 410–418. [Google Scholar] [CrossRef]
  19. Herdt, A.; Brown, R.D.; Scott-Fleming, I.; Cao, G.; MacDonald, M.; Henderson, D.; Vanos, J.K. Outdoor Thermal Comfort during Anomalous Heat at the 2015 Pan American Games in Toronto, Canada. Atmosphere 2018, 9, 321. [Google Scholar] [CrossRef]
  20. Spagnolo, J.; de Dear, R. A field study of thermal comfort in outdoor and semioutdoor environments in subtropical Sydney Australia. Build. Environ. 2003, 38, 721–738. [Google Scholar] [CrossRef]
  21. Hwang, R.L.; Lin, T.P. Thermal comfort requirements for occupants of semi-outdoor and outdoor environments in hot-humid regions. Archit. Sci. Rev. 2007, 50, 357–364. [Google Scholar] [CrossRef]
  22. Papachristou, C.; Foteinaki, K.; Kazanci, O.B.; Olesen, B.W. Structures that Include a Semi-Outdoor Space: Part 2: Thermal Environment. In Proceedings of the 12th REHVA World Congress, Aalborg, Denmark, 22–25 May 2016. [Google Scholar]
  23. Brazel, A.J.; Marcus, M.G. Heat Enhancement by Longwave Wall Emittance. Geogr. Rev. 1987, 4, 440–455. [Google Scholar] [CrossRef]
  24. Gutter, B. Assessing the Microclimate of Bryant-Denny Stadium. J. Sci. Health Univ. Ala. (JOSHUA) 2011, 8, 14–18. [Google Scholar]
  25. Reddick, T.R.; Vanos, J.K. A new approach to monitor and map heat exposure in a semi-outdoor environment: A football stadium case study in west Texas. In Proceedings of the 15th Annual Student Conference, New Orleans, LA, USA, 9 January 2016; American Meteorological Society: Boston, MA, USA, 2016. Available online: https://ams.confex.com/ams/96Annual/webprogram/Paper292220.html (accessed on 31 July 2024).
  26. Bouyer, J.; Vinet, J.; Delpech, P.; Carré, S. Thermal comfort assessment in semioutdoor environments: Application to comfort study in stadia. J. Wind Eng. Ind. Aerodyn. 2007, 95, 963–976. [Google Scholar] [CrossRef]
  27. Fiala, D.; Lomas, K.J. Applications of a computer model predicting human thermal comfort and responses to the design of sports stadia. In CIBSE; Institute of Energy and Sustainable Design, De Montfort University: Leicester, UK, 1999; p. 492. [Google Scholar]
  28. Helbing, D.; Johansson, A. Pedestrian, Crowd, and Evacuation Dynamics. Encycl. Complex. Syst. Sci. 2013, 16, 6476–6495. [Google Scholar] [CrossRef]
  29. Stewart, I.D.; Kennedy, C.A. Metabolic heat production by human and animal populations in cities. Int. J. Biometeorol. 2017, 61, 1159–1171. [Google Scholar] [CrossRef]
  30. Davis Wade Stadium at Scott Field. Available online: https://express.adobe.com/page/bz0wAZIToDGjz/ (accessed on 21 July 2024).
  31. Habeeb, D.; Clawson, J.; Zakeresfahani, A.; Holtz, Z. Investigating and validating on-body temperature sensors for personal heat exposure tracking. In Proceedings of the CHI Conference on Human Factors in Computing Systems, Orleans, LA, USA, 29 April–5 May 2022; Volume 343, pp. 1–14. [Google Scholar]
  32. Schaefer, G.L.; Cosh, M.H.; Jackson, T.J. The USDA Natural Resources Conservation Service Soil Climate Analysis Network (SCAN). J. Atmos. Ocean. Technol. 2007, 24, 2073–2077. [Google Scholar] [CrossRef]
  33. Bailey, E.; Fuhrmann, C.M.; Runkle, J.; Stevens, S.; Brown, M.E.; Sugg, M.M. Wearable sensors for personal temperature exposure assessments: A comparative study. Environ. Res. 2020, 180, 108858. [Google Scholar] [CrossRef]
  34. Scott, A.A.; Zaitchik, B.; Waugh, D.W.; O’Meara, K. Intraurban temperature variability in Baltimore. J. Appl. Meteorol. Climatol. 2017, 56, 159–171. [Google Scholar] [CrossRef]
  35. Wen, C.; Mamtimin, A.; Feng, J.; Wang, Y.; Yang, F.; Huo, W.; Zhou, C.; Li, R.; Song, M.; Gao, J.; et al. Diurnal Variation in Urban Heat Island Intensity in Birmingham: The Relationship between Nocturnal Surface and Canopy Heat Islands. Land 2023, 12, 2062. [Google Scholar] [CrossRef]
  36. Alfano, F.R.; Palella, B.I.; Riccio, G. Thermal environment assessment reliability using temperature—Humidity indices. Ind. Health 2011, 49, 95–106. [Google Scholar] [CrossRef] [PubMed]
  37. de Freitas, C.R.; Grigorieva, E.A. A comprehensive catalogue and classification of human thermal climate indices. Int. J. Biometeorol. 2015, 59, 109–120. [Google Scholar] [CrossRef] [PubMed]
  38. Rothfusz, L.P. The Heat Index “Equation” (or, More than You Ever Wanted to Know about Heat Index); NWS Southern Region Technical Attachment 1990, SR/SSD 90-23: Fort Worth, TX, USA, 1990; p. 2. Available online: https://www.weather.gov/media/bgm/ta_htindx.PDF (accessed on 31 July 2024).
  39. Reza, M.; Daneshvar, M.; Bagherzadeh, A.; Tavousi, T. Assessment of Bioclimatic Comfort Conditions based on Physiologically Equivalent Temperature (PET) using the RayMan Model in Iran. Cent. Eur. J. Geosci. 2013, 5, 53–60. [Google Scholar]
  40. Matzarakis, A.; Fröhlich, D. Sport events and climate for visitors—The case of FIFA World Cup in Qatar 2022. Int. J. Biometeorol. 2015, 59, 481–486. [Google Scholar] [CrossRef]
  41. Hoppe, P. The physiological equivalent temperature—A universal index for the biometeorological assessment of the thermal environment. Int. J. Biometeorol. 1999, 43, 71–75. [Google Scholar] [CrossRef]
  42. Matzarakis, A.; Amelung, B. Physiologically equivalent temperature as indicator for impacts of climate change on thermal comfort of humans. In Seasonal Forecasts, Climatic Change and Human Health; Thomson, M.C., Garcia-Herrera, R., Beniston, M., Eds.; Springer: Dordrecht, The Netherlands, 2008; pp. 161–172. [Google Scholar]
  43. Matzarakis, A.; Rutz, F.; Mayer, H. Modelling radiation fluxes in simple and complex environments: Basics of the RayMan model. Int. J. Biometeorol. 2007, 54, 131–139. [Google Scholar] [CrossRef]
  44. Mei, W.; Qu, M. Evaluation and Analysis of Wind Flow for a Football Stadium. Procedia Eng. 2016, 145, 774–781. [Google Scholar] [CrossRef]
  45. Chen, Y.-C.; Matzarakis, A. Modified physiologically equivalent temperature—Basics and applications for western Europe climate. Theor. Appl. Climatol. 2018, 132, 1275–1289. [Google Scholar] [CrossRef]
  46. Dzyuban, Y.; Hondula, D.; Vanos, J.; Middel, A.; Coseo, P.; Kuras, E.; Redman, C. Evidence of alliesthesia during a neighborhood thermal walk in a hot and dry city. Sci. Total Environ. 2022, 834, 155294. [Google Scholar] [CrossRef]
  47. Ainsworth, B.E.; Haskell, W.L.; Hermann, S.D.; Meckes, N.; Bassett, D.R., Jr.; Tudor-Locke, C.; Greer, J.L.; Vezina, J.; Whitt-Glover, M.C.; Leon, A.S. Compendium of physical activities: A second update of codes and MET values. Med. Sci. Sports Exerc. 2011, 43, 1575–1581. [Google Scholar] [CrossRef]
  48. Deb, C.; Ramachandraiah, A. The significance of Physiological Equivalent Temperature (PET) in outdoor thermal comfort studies. Int. J. Eng. Sci. Technol. 2010, 2, 2825–2828. [Google Scholar]
  49. Rupp, R.F.; Vásquez, N.G.; Lamberts, R. A review of human thermal comfort in the built environment. Energy Build. 2015, 105, 178–205. [Google Scholar] [CrossRef]
  50. Taleghani, M.; Kleerekoper, L.; Tenpierik, M.; Dobbelsteen, A. Outdoor thermal comfort within five different urban forms in the Netherlands. Build. Environ. 2015, 83, 65–78. [Google Scholar] [CrossRef]
  51. Martinelli, L.; Lin, T.; Matzarakis, A. Assessment of the influence of daily shadings pattern on human thermal comfort and attendance in Rome during summer period. Build. Environ. 2015, 92, 30–38. [Google Scholar] [CrossRef]
  52. Matzarakis, A.; Mayer, H. Heat stress in Greece. Int. J. Biometeorol. 1997, 41, 34–39. [Google Scholar] [CrossRef]
  53. Matzarakis, A.; Moses, H.M.; Iziomon, G. Applications of a universal thermal index: Physiological equivalent temperature. Int J. Biometeorol. 1999, 43, 76–84. [Google Scholar] [CrossRef]
  54. Jackson NWS Product Guide for Advisory and Warning Criteria. Available online: https://www.weather.gov/jan/productguide_nonprecip (accessed on 31 July 2024).
  55. Lu, Y.-C.; Romps, D.M. Extending the heat index. J. Appl. Meteorol. Climatol. 2022, 61, 1367–1383. [Google Scholar] [CrossRef]
  56. Klemm, W.; Heusinkveld, B.G.; Lenzholzer, S.; Jacobs, M.H.; Van Hove, B. Psychological and physical impact of urban green spaces on outdoor thermal comfort during summertime in The Netherlands. Build. Environ. 2015, 83, 120–128. [Google Scholar] [CrossRef]
Atmosphere 15 01184 g001
Figure 1. Map of the location of Starkville, MS, USA (star; upper left) and aerial image of Davis Wade Stadium obtained from Google Earth (center).
Figure 1. Map of the location of Starkville, MS, USA (star; upper left) and aerial image of Davis Wade Stadium obtained from Google Earth (center).
Atmosphere 15 01184 g001
Atmosphere 15 01184 g002
Figure 2. Photograph of an iButton from Davis Wade Stadium mounted in a plastic fob and attached to a railing with a zip tie.
Figure 2. Photograph of an iButton from Davis Wade Stadium mounted in a plastic fob and attached to a railing with a zip tie.
Atmosphere 15 01184 g002
Atmosphere 15 01184 g003
Figure 3. (A) Locations of Thermochrons placed in the seating areas (circles) and concourses (triangles). (B) Locations of Hygrochrons placed in the seating areas, at field level, and in the concourses (numbered boxes). Stars denote the locations of photographs taken for sky view factor calculations.
Figure 3. (A) Locations of Thermochrons placed in the seating areas (circles) and concourses (triangles). (B) Locations of Hygrochrons placed in the seating areas, at field level, and in the concourses (numbered boxes). Stars denote the locations of photographs taken for sky view factor calculations.
Atmosphere 15 01184 g003
Atmosphere 15 01184 g004
Figure 4. Plot of average hourly iButton temperatures (°F) recorded within the seating and field sections of the stadium on 29 October 2016. The first map (A) shows temperatures one hour prior to the game start time (1:00 p.m.). Maps (B–E) show temperatures from the game start time to the game end time (2:00 p.m. to 5:00 p.m.). The last map (F) shows temperatures one hour after the game’s end time (6:00 p.m.). Values displayed on each map are for individual iButtons. Inverse distance weighting interpolation was used to create the contour maps in ArcGIS. Contour lines and shading are in 1°F increments ranging from 81 °F (dark green) to 109 °F (pink). High-resolution versions of each map can be found in Supplemental File.
Figure 4. Plot of average hourly iButton temperatures (°F) recorded within the seating and field sections of the stadium on 29 October 2016. The first map (A) shows temperatures one hour prior to the game start time (1:00 p.m.). Maps (B–E) show temperatures from the game start time to the game end time (2:00 p.m. to 5:00 p.m.). The last map (F) shows temperatures one hour after the game’s end time (6:00 p.m.). Values displayed on each map are for individual iButtons. Inverse distance weighting interpolation was used to create the contour maps in ArcGIS. Contour lines and shading are in 1°F increments ranging from 81 °F (dark green) to 109 °F (pink). High-resolution versions of each map can be found in Supplemental File.
Atmosphere 15 01184 g004
Atmosphere 15 01184 g005
Figure 5. Same as Figure 4, except for temperatures on 10 September 2016. The first map (A) shows temperatures one hour prior to the game start time (5:00 p.m.). Maps (B–E) show temperatures from the game start time to the game end time (6:00 p.m. to 9:00 p.m.). The last map (F) shows temperatures one hour after the game’s end time (10:00 p.m.). Contour lines and shading are in 1 °F increments ranging from 76 °F (dark green) to 107 °F (pink). High-resolution versions of each map can be found in Supplemental File.
Figure 5. Same as Figure 4, except for temperatures on 10 September 2016. The first map (A) shows temperatures one hour prior to the game start time (5:00 p.m.). Maps (B–E) show temperatures from the game start time to the game end time (6:00 p.m. to 9:00 p.m.). The last map (F) shows temperatures one hour after the game’s end time (10:00 p.m.). Contour lines and shading are in 1 °F increments ranging from 76 °F (dark green) to 107 °F (pink). High-resolution versions of each map can be found in Supplemental File.
Atmosphere 15 01184 g005
Atmosphere 15 01184 g006
Figure 6. Hourly heat index values (°F) recorded within the stadium and at the SCAN station beginning one hour prior and ending one hour after the end of the game on (A) 3 September 2016, (B) 10 September 2016, and (C) 29 October 2016. Column numbers correspond with the numbered locations in Figure 3B. Values in the chart are color-coded according to the National Weather Service heat stress categories provided in Table 2.
Figure 6. Hourly heat index values (°F) recorded within the stadium and at the SCAN station beginning one hour prior and ending one hour after the end of the game on (A) 3 September 2016, (B) 10 September 2016, and (C) 29 October 2016. Column numbers correspond with the numbered locations in Figure 3B. Values in the chart are color-coded according to the National Weather Service heat stress categories provided in Table 2.
Atmosphere 15 01184 g006
Atmosphere 15 01184 g007
Figure 7. Examples of graphics illustrating the (a) temperature and (b) heat index in different parts of the stadium and how they compare to conditions outside the stadium (i.e., at the nearest airport). Graphics created by Joel Young (WTVA, Tupelo, MS, USA).
Figure 7. Examples of graphics illustrating the (a) temperature and (b) heat index in different parts of the stadium and how they compare to conditions outside the stadium (i.e., at the nearest airport). Graphics created by Joel Young (WTVA, Tupelo, MS, USA).
Atmosphere 15 01184 g007
Table 1. Summary of bootstrap results of the mean and standard deviation (with 95% confidence intervals) of temperature and heat index (°C) for each region of the stadium and the weather station (SCAN) during the study period. Values in italics are statistically different from the seating and field areas (p < 0.05). Values in bold and italics are statistically different from all other locations (p < 0.05).
Table 1. Summary of bootstrap results of the mean and standard deviation (with 95% confidence intervals) of temperature and heat index (°C) for each region of the stadium and the weather station (SCAN) during the study period. Values in italics are statistically different from the seating and field areas (p < 0.05). Values in bold and italics are statistically different from all other locations (p < 0.05).
Mean TemperatureStandard Deviation
2.5%50%97.5%2.5%50%97.5%
Seating21.922.322.66.97.17.3
Field21.722.022.36.87.07.2
Concourse22.522.722.95.55.75.9
SCAN20.520.821.27.37.57.7
Mean Heat IndexStandard Deviation
2.5%50%97.5%2.5%50%97.5%
Seating31.732.032.33.53.63.8
Field31.331.531.83.03.13.3
Concourse30.230.430.61.81.92.1
SCAN31.031.331.53.03.13.2
Table 2. Heat stress alert categories based on heat index (HI) and physiological equivalent temperature (PET). HI categories adopted from the U.S. National Weather Service. PET categories adopted from [52]. Temperatures provided in Fahrenheit and Celsius.
Table 2. Heat stress alert categories based on heat index (HI) and physiological equivalent temperature (PET). HI categories adopted from the U.S. National Weather Service. PET categories adopted from [52]. Temperatures provided in Fahrenheit and Celsius.
HI CategoryHI Range (°C)PET CategoryPET Range (°C)Effects on the Body Based on HI (PET)
Comfortable64–74
(18–23)		No thermal stress
Caution80–90
(27–32)		Slightly Warm to Warm73–95
(23–35)		Fatigue possible with prolonged exposure and/or physical activity (slight to moderate heat stress)
Extreme Caution91–103
(32–41)		Hot95–106
(35–41)		Heat stroke, heat cramps, or heat exhaustion possible with prolonged exposure and/or physical activity (strong heat stress)
Danger104–125
(41–54)		Very Hot>106
(41)		Heat cramps or heat exhaustion likely, and heat stroke possible with prolonged exposure and/or physical activity (extreme heat stress)
Extreme Danger>125
(54)		 Heat stroke likely
Table 3. Summary of bootstrap results of the mean and standard deviation (with 95% confidence intervals) of mPET (°C) under varying wind speed thresholds for each region of the stadium and the weather station (SCAN) during the study period. Values in bold and italics are statistically different from all other locations (p < 0.05).
Table 3. Summary of bootstrap results of the mean and standard deviation (with 95% confidence intervals) of mPET (°C) under varying wind speed thresholds for each region of the stadium and the weather station (SCAN) during the study period. Values in bold and italics are statistically different from all other locations (p < 0.05).
Mean mPETStandard Deviation
2.5%50%97.5%2.5%50%97.5%
Seating36.436.937.48.28.58.7
Field35.335.836.37.78.08.2
Concourse33.834.234.77.47.67.8
SCAN35.235.736.27.57.78.0
Mean mPETStandard Deviation
2.5%50%97.5%2.5%50%97.5%
Seating32.132.633.15.96.26.4
Field31.732.232.65.65.86.0
Concourse29.830.230.64.85.05.2
SCAN31.131.331.95.45.65.8
Mean mPETStandard Deviation
2.5%50%97.5%2.5%50%97.5%
Seating31.131.632.05.45.75.9
Field31.031.532.05.05.35.5
Concourse29.730.230.64.14.34.5
SCAN30.330.731.14.95.15.3
Mean mPETStandard Deviation
2.5%50%97.5%2.5%50%97.5%
Seating30.731.231.65.25.55.7
Field30.330.831.34.85.15.3
Concourse28.629.029.43.63.84.0
SCAN29.930.230.54.74.95.1
Table 4. Summary of first aid and EMS heat-related cases and corresponding HI and mPET values (°C) for the east and west sides of the seating area during the three games identified as thermally oppressive.
Table 4. Summary of first aid and EMS heat-related cases and corresponding HI and mPET values (°C) for the east and west sides of the seating area during the three games identified as thermally oppressive.
3 September 201610 September 201629 October 2016
WestEastWestEastWestEast
Heat-related cases per 10,000 spectators1.571.931.211.380.340.69
Mean HI39.642.634.134.331.335.2
Mean mPET Wind 0.150.951.727.528.139.141.4
Mean mPET Wind 6.140.942.126.127.231.835.5
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Collins, A.; Brown, M.; Gutter, B.; Fuhrmann, C. Microclimatic Variability and Thermal Comfort of Spectators in an Outdoor Stadium Venue. Atmosphere 2024, 15, 1184. https://doi.org/10.3390/atmos15101184

AMA Style

Collins A, Brown M, Gutter B, Fuhrmann C. Microclimatic Variability and Thermal Comfort of Spectators in an Outdoor Stadium Venue. Atmosphere. 2024; 15(10):1184. https://doi.org/10.3390/atmos15101184

Chicago/Turabian Style

Collins, Andrew, Michael Brown, Barrett Gutter, and Christopher Fuhrmann. 2024. "Microclimatic Variability and Thermal Comfort of Spectators in an Outdoor Stadium Venue" Atmosphere 15, no. 10: 1184. https://doi.org/10.3390/atmos15101184

APA Style

Collins, A., Brown, M., Gutter, B., & Fuhrmann, C. (2024). Microclimatic Variability and Thermal Comfort of Spectators in an Outdoor Stadium Venue. Atmosphere, 15(10), 1184. https://doi.org/10.3390/atmos15101184

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop