Wet bulb globe temperature from climate model outputs: a method for projecting hourly site-specific values and trends.

Wet bulb globe temperature from climate model outputs: a method for projecting hourly site-specific values and trends. | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Wet bulb globe temperature from climate model outputs: a method for projecting hourly site-specific values and trends. Erik Patton, Wenhong Li, Ashley Ward, Martin Doyle This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4414813/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 17 Sep, 2024 Read the published version in International Journal of Biometeorology → Version 1 posted 4 You are reading this latest preprint version Abstract Increasing temperature will impact future outdoor worker safety but quantifying this impact to develop local adaptations is challenging. Wet bulb globe temperature (WBGT) is the preferred thermal index for regulating outdoor activities in occupational health, athletic, and military settings, but global circulation models (GCMs) have coarse spatiotemporal resolution and do not always provide outputs required to project the full diurnal range of WBGT. This article presents a novel method to project WBGT at local spatial and hourly temporal resolutions without many assumptions inherent in previous research. We calculate sub-daily future WBGT from GCM output and then estimate hourly WBGT based on a site-specific, historical diurnal cycles. We test this method against observations at U.S. Army installations and find results match closely. We then project hourly WBGT at these locations from January 1, 2025, to December 31, 2100, to quantify trends and estimate future periods exceeding outdoor activity modification thresholds. We find regional patterns affecting WBGT, suggesting accurately projecting WBGT demands a localized approach. Results show increased frequency of hours at high WBGT and, using U.S. military heat thresholds, we estimate impacts to future outdoor labor. By mid-century, some locations are projected to experience an average of 20 or more days each summer when outdoor labor will be significantly impacted. The method’s fine spatiotemporal resolution enables detailed analysis of WBGT projections, making it useful applied at specific locations of interest. Climate change wet bulb globe temperature heat illness outdoor labor Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Climate change and associated increasing global temperatures (IPCC 2021 ) will affect many aspects of society, including outdoor labor. However, local impacts of climate change and warming depend on regional geographies, weather patterns, and local conditions. Understanding the local impact of climate change on outdoor labor thus demands local scale analysis. Additionally, temperature is rarely the best index to measure thermal influences on human health (Li, Zhang et al. 2017 , Buzan and Huber 2020 , Pisaniello and Di Corleto 2023). Factors such as humidity, wind, sun exposure, and exertional effort are relevant for outdoor workers. To consider these variables on thermal stress, over 100 variants of temperature indices have been developed (Blazejczyk, Epstein et al. 2012 ). Of these, the wet bulb globe temperature (WBGT) index is widely used to establish activity modification thresholds to prevent heat illness (Budd 2008 , Kong and Huber 2022 ). For example, WBGT is used to regulate athletic practice and competition (Roberts, Armstrong et al. 2021 , Yeargin, Hirschhorn et al. 2023 ), establish occupational health and safety standards (Parsons 2006 , Périard, DeGroot et al. 2022 ), and modify training in military settings (Minard 1961 , HQDA 2022 ). Over most of the northern hemisphere, new high summer mean WBGT records are projected to occur more frequently than new mean high records measured by surface temperature alone (Li, Zhang et al. 2017 ). WBGT is the weighted average of the natural wet bulb temperature ( T wb ), dry bulb temperature ( T a ), and black globe temperature ( T bg ) (Eq. 1). Each of these WBGT variables respond to weather variables differently. T a is measured by a shaded thermometer. T wb cools by latent heat loss and is a proxy for sweating, which removes heat from the skin during sweat evaporation. Latent heat loss is influenced by humidity, with less evaporation (and less heat removed) during periods of high relative humidity. T bg is influenced by radiant heat sources, primarily solar radiation, and is measured inside a 6-inch black metal sphere (Budd 2008 ). Equation 1: Wet Bulb Globe Temperature $$WBGT Index=0.7\left({T}_{wb}\right)+0.2\left({T}_{bg}\right)+0.1\left({T}_{a}\right)$$ The WBGT index is often used to establish safety and activity modification thresholds for populations at risk of exertional heat illness (Hosokawa, Casa et al. 2019 ). During physical labor skeletal muscles generate large heat loads (Sawka, Leon et al. 2011 ) which, if not dissipated, increase the body’s core temperature. This can lead to exertional heat illness (EHI) ranging from heat cramps to potentially fatal heat stroke (Alele, Malau-Aduli et al. 2020 ). When ambient temperature exceeds skin temperature of ~ 35 o C, sweat evaporation becomes the only heat dissipation method possible (Sawka, Leon et al. 2011 ). Heat loss from sweat evaporation decreases with increasing relative humidity until no evaporation, and therefore no net heat loss, which occurs at relative humidity of 100%. Heat stress is thus maximized when temperature and humidity are high and is additionally compounded by radiant solar heat when outdoors. Safety thresholds use WBGT to protect against EHI are often calibrated for specific populations, such as the guidelines set forth by the American College of Sports Medicine (ACSM) for athletes, worker safety recommendations from the Occupational Safety and Health Administration (OSHA), and work-rest cycles for military personnel set by the U.S. Department of the Army. The ACSM recommends activity modification when WBGT exceeds values as low as 15.1 o C and competition cancellation when WBGT exceeds values between 24.6–32.3 o C (Roberts, Armstrong et al. 2021 ) while the U.S. military modifies some outdoor training beginning at a WBGT of 25.56 o C (HQDA 2022 ). This study uses values from the military scale when assessing impacts to outdoor labor and training (Table 1 ). Table 1. WBGT index thresholds used to guide activity modification by the U.S. Military. WBGT index, o F WBGT index, o C Heat Category Flag Color minimum maximum minimum maximum 1 White 78 81.9 25.56 27.77 2 Green 82 84.9 27.78 29.43 3 Yellow 85 87.9 29.44 31.10 4 Red 88 89.9 31.11 32.21 5 Black 90 - 32.22 - WBGT can be accurately calculated from past weather observations (Liljegren, Carhart et al. 2008 , Lemke and Kjellstrom 2012 , Patel, Mullen et al. 2013 , Kong and Huber 2022 ), but limitations in the spatiotemporal resolution of global circulation models (GCMs) and the large computational effort needed to explicitly calculate WBGT from meteorological observations in large data sets (Brimicombe, Lo et al. 2023 ) complicate projecting future WBGT accurately. Explicitly calculating WBGT requires five weather variables and iterative calculations to resolve both \({T}_{bg}\) and \({T}_{wb}\) . GCMs typically provide such variables as daily mean values, but greater temporal resolution is needed to estimate local hourly impacts. In addition, WBGT is often sensitive to small scale local features, but GCM output is provided at coarse spatial resolution. Spatial resolution challenges can be partially mitigated by using downscaled GCM output, but the highest resolution data sets with all required variables is the 0.25 x 0.25 degree (approximately 25km by 25km) output from the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) (Thrasher 2021), which may still mask the impact of local conditions on projected WBGT. Previous studies commonly address these challenges by omitting the \({T}_{bg}\) and considering regional or global trends (Willett and Sherwood 2012 , Knutson and Ploshay 2016 , Li, Zhang et al. 2017 , Parsons, Shindell et al. 2021 ). Omitting \({T}_{bg}\) is used to estimate indoor WBGT and simplify calculations but can underestimate outdoor WBGT. Methods that include \({T}_{bg}\) exist and have been applied to global gridded data sets (Brimicombe, Lo et al. 2023 ) yet may not capture local variations due to coarse spatial resolution, failing to adequately project WBGT at a specific location of interest. Takakura et al ( 2019 ) construct site-specific WBGT using a method conceptually similar to ours but simplify the computational effort by applying regressions to estimate WBGT from GCM outputs instead of explicitly calculating it; our method appears more accurate at recreating historical WBGT observations, and this additional accuracy is presumably carried forward when estimating future WBGT values. This study describes a novel method for providing site-specific WBGT projections based on historical observations and downscaled GCM output. Our key contribution is in removing many assumptions required in previous work by explicitly calculating WBGT, including the \({T}_{bg}\) component, doing so at a local scale. We do this by evaluating the daily maximum, mean, and minimum WBGT on future days using downscaled GCM outputs. We make assumptions only for variables not included in the downscaled GCM data sets (e.g., surface pressure); these assumptions are constrained by historical observations. Projected WBGT values on intervening hours are then interpolated using location-specific historical climatological averages. From an initial large ensemble of GCMs, a subset of models selected for skill in recreating historical extreme value observations is then bias adjusted using historical observations to ensure a site-specific final output. We test out method against records from four U.S. army training installations: Ft Jackson, South Carolina (FJSC); Ft Moore, Georgia (FMGA); Ft Leonard Wood, Missouri (FLW); and Ft Sill, Oklahoma (FSOK) (Fig S1 ). Choosing these locations allows us to compare historical WBGT records provided by the U.S Air Force 14th Weather Squadron (14th WX) with WBGT calculated from our methodology. We then apply our method using GCMs selected for skill at each location and bias-adjust the output using historical records to project hourly WBGT. Lastly, we analyze our WBGT projections to determine how future trends and extremes may affect outdoor training time at these locations. Methods Data Sources Airfield weather stations at or near the study locations (Table S2) provide our hourly weather records for 01/01/1990 to 09/30/2022 including surface air temperature ( T a in o C), surface pressure (in hPa), relative humidity (in g/kg), solar radiation (in W/m 2 ), and 10m wind speed (in m/s). Missing values are interpolated as described in Patton and Doyle ( 2023 ). Continuous fixed point WBGT observations are rare (Takakura, Fujimori et al. 2019 ) and, to our knowledge, no multi-decade data sets exist. Instead, we use hourly WBGT records for these locations obtained from the 14th WX, the organization providing authoritative weather and climate data to the U.S. military (HQDA 2021 ), as historical observations. GCM data sets are obtained from the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) (Thrasher 2021). We use the single grid cell at each location encompassing the weather station providing our historical records. Our methodology requires five input variables: temperature ( T a ), relative humidity, wind speed, solar radiation, and surface pressure. Of 35 GCMs included in the NEX-GDDP CMIP6 project, 23 include all required variables. After dropping one GCM known to have temperature deviations (Thrasher 2023 ) and three due to data access difficulties, we subject 19 to full analysis at each location (Table S3). Data sets are subdivided into three periods: 2000–2014 (15 years) using GCM historical runs to evaluate the accuracy of our method against observations; 2015–2021 (7 years) using GCM scenario runs compared against observations for bias adjustment; and 2025–2100 (76 years) to project future WBGT at study locations. We select the two IPCC scenario Shared Socioeconomic Pathways (SSPs) that currently appear to reflect more plausible future outcomes (Hausfather and Peters 2020 ): SSP 2-4.5 middle-of-the-road and SSP 3–7.0 regional rivalry (Hausfather 2018 ). SSP 3 is becoming the choice scenario for planning in a high emission future in some impact assessments (NMFS 2023 ) and SSP 2 the most likely scenario given globally pledged climate policies (Scafetta 2024 ). In contrast, SSP 1 and SSP 5 scenarios appear increasingly unlikely (Kriegler, Bauer et al. 2017 , Riahi, van Vuuren et al. 2017 ). SSP 2 and SSP 3 also bracket the likely range of warming (2.1-3.4 o C, median 2.8 o C) projected by 2100 (IPCC 2023 ). This likely warming range falls within the projected range for SSP 2 (2.1-3.5 o C). SSP 3’s higher range (2.8-4.6 o C) makes it suitable for lower-probability but higher-risk impact evaluation. Evaluating the Liljegren method We calculate WBGT as described by Liljegren (2008) using R package “wbgt” (Lieblich 2017 ). This method is recommended in comparative studies (Lemke and Kjellstrom 2012 , Kong and Huber 2022 ) and implemented in similar work (Ahn, Uejio et al. 2022 , Lewandowski, Kioumourtzoglou et al. 2022 ). We evaluate the method in three ways: (1) using hourly historical weather observations (2000–2014) as inputs and comparing results to WBGT provided by the 14th WX, (2) correlating weather observations with daily maximum, mean, and minimum WBGT from the 14th WX and Liljegren methods, and (3) using steps #1–3 of this study’s methodology (described below) to estimate WBGT over the same period using inputs from 19 GCM historical runs. Finding the Liljegren method appropriate, we apply our novel methodology to estimate hourly WBGT at each location between 01/01/2025 and 12/31/2100. We use 5 steps: (1) finding hourly WBGT for a climatological average year, (2) estimating required weather observation values unavailable from GCM outputs, (3) creating uncorrected hourly WBGT by shifting the average WBGT diurnal curve to match WBGT values calculated from GCM outputs, (4) selecting a six model GCM subset based on a goodness-of-fit test evaluating model skill, and (5) bias adjusting the subset. Step #1: WBGT values in an “average year” Our first step is finding mean values at each hour of a year for T a , wind speed, relative humidity, solar radiation, and surface pressure using observations between 01/01/1990–09/30/2022 and after removing leap days. The resulting data set contains 1 years’ worth of mean hourly observations ( n = 8,760) for each variable. We consider these the variable climatological averages for their respective hour of the year. The Liljegren method can return unrealistic values if wind ≅ 0 m/s. Previous WBGT studies used minimums of 0.5 m/s (Spangler, Liang et al. 2022 ) or 1.0 m/s (Lemke and Kjellstrom 2012 ), or apply a correction factor to velocities below 0.5 m/s (Patel, Mullen et al. 2013 ), recognizing in “still air” body movement and natural processes generate some minimum airflow over skin. GCM wind speed output is provided at 10m height, so we estimate a speed of 0.62 m/s returning 0.5 m/s at 2m (Eq. 2) and adjust all observations < 0.62 m/s to 0.62 m/s. Equation 2: Windspeed height profile formula \({v}_{2}={v}_{1}*\left[\right(\text{ln}\left(\frac{{h}_{2}}{{z}_{0}}\right))/(\text{ln}\left(\frac{{h}_{1}}{{z}_{0}}\right))\) ] Where: \({v}_{x}\) and \({h}_{x}\) are the velocity (in m/s) of wind at height h \({z}_{0}\) is our chosen roughness length of 0.0024m After adjusting wind speed, hourly WBGT is estimated for every hour between 01/01/1990–09/30/2022. The mean WBGT for each hour across years is then calculated, establishing a climatological average WBGT value for each hour of the year. This climatological average WBGT, consisting of 365 averaged WBGT diurnal cycles, forms the baseline from which future hourly values are later derived. Step #2: Creating input variables: the Liljegren method and the NEX-GDDP data sets. Here we describe how the required input variables are obtained or estimated from GCM output. Our method requires daily maximum, mean, and minimum values for each input variable. The NEX-GDDP data sets provide six: daily max, mean, and min T a , and daily mean wind speed, relative humidity, and solar radiation. Many NEX-GDDP GCM outputs do not include surface pressure, leaving nine missing values to be estimated: daily maximum and minimum for wind speed, solar radiation, and relative humidity, and daily maximum, mean, and minimum surface pressure. Relative humidity responds non-linearly with changing T a , making it difficult to estimate maximum and minimum from a daily mean value. We choose instead to calculate relative humidity values from specific humidity (described below). To ensure projected WBGTs are mapped to the correct part of the diurnal cycle, the hour associated with minimum, mean, and maximum WBGT at each location and on each day of the year is identified from the 14th WX data set. This allows the time associated with maximum and minimum WBGT to change depending on seasonal and local considerations. 2.1 Estimating wind speed and solar radiation To determine maximum and minimum wind speed we use a scaling factor unique to each location and day of the year. We find speed corresponding with the hour of daily maximum, mean, and minimum WBGT across fifteen years of observations (01/01/2000–12/31/2014). The difference between max (min) and mean wind speed is calculated for each day. Differences are grouped by day to create a unique average daily scaling factor for mean-to-max and mean-to-min wind speed for each day and location. The same method is applied to estimate daily maximum solar radiation. Minimum solar radiation is fixed at 0 W/m 2 on the assumption that minimum WBGT occurs at night. To prevent outliers from skewing scaling factors in this relatively small sample (the mean of 15 observations creates each daily scaling factor), observations > 99th and < 1st percentile are replaced with the 1st and 99th percentile. The result is three sets of 365 scaling factor values for each location: one each for the difference between maximum-to-mean and minimum-to-mean wind speed and one between maximum-to-mean solar radiation. Scaling factors are added to corresponding daily mean GCM values to estimate daily maximum and minimum values, inherently accounting for intra-year and seasonal changes. We correct the few unrealistic maximum solar radiation values by limiting solar radiation to each locations historical observed maximum. Daily maximum wind speed values greater than those in our historical data sets, also rare, are retained on the assumption that GCMs may predict future wind speeds above those previously recorded (Table S4). 2.2 Estimating surface pressure. Although similar studies have used the ideal gas law to estimate surface pressure based on temperature and elevation (Chavaillaz, Roy et al. 2019 ) our study found it performs poorly when compared against hourly observations. Instead, three linear models are developed at each study location using historical records (Eq. 3). These models correlate observed surface pressure at hours associated with WBGT maximum, mean, and minimum with the T a , wind speed, relative humidity, and solar radiation observed at those same hours to estimate surface pressure distributions better matching observations. Equation 3: Surface Pressure Linear Model $$Surface Pressure ={}_{1}\left({T}_{a}\right)+{}_{2}\left(Wind Speed\right)+{}_{3}\left(Relative Humidity\right)+{}_{4}\left(Solar Radition\right)+X$$ To remove any systemic bias in the distribution of maximum or minimum surface pressure, values are shifted by the difference between the mean observed and calculated maximum (minimum) surface pressure values, “nudging” the distribution to align more closely with observations (Fig. S5). These surface pressure values are location specific since models are developed from local weather records. 2.3 Estimating relative humidity. We use daily mean specific humidity to calculate relative humidity. By assuming the mass of water vapor remains constant throughout the day, daily minimum, mean, and maximum relative humidity can be found as a function of temperature and pressure. We implement this calculation with R package ‘humidity’ (Cai 2019 ). Inputs are daily mean specific humidity and daily max, mean, and minimum T a (from GCMs), and daily max, mean, and min surface pressure. Specific humidity was not provided in the 14th WX data, preventing direct comparison between observed and calculated relative humidity using this method. As a proxy to test robustness of this method, relative humidity values from historical GCM runs are calculated and compared to observations between 2000–2014. Calculated values are slightly overestimated at the hour of daily minimum WBGT, while distributions better match observations at mean and maximum WBGT. (Fig. S6). Step #3: Uncorrected Hourly Wet Bulb Globe Temperatures After obtaining all 15 daily input values, we calculate daily maximum, mean, and minimum WBGT per Liljegren (2008). To extend WBGT to hourly values, we shift the climatological average diurnal curves from step #1 to new positions by calculating the difference between the GCM-derived WBGT and the corresponding WBGT from the climatological average year. For example, GCM-derived WBGTs of 25 o C (maximum), 20 o C (mean), and 15 o C (minimum) on May 1st, 2050, and climatological average WBGT on May 1st of 23 o C, 20 o C, and 12 o C, result in differences associated with May 1st, 2050, of 2 o C, 0 o C, and 3 o C (Table S7). Since diurnal cycles form sinusoidal waves, mean differences occur twice daily. Each day’s differences are therefore associated with four hours- once at maximum WBGT, once at minimum, and twice at mean. For the remaining twenty hours the difference between GCM WBGT and climatological average WBGT is created by interpolating between these four differences. All differences are then added back to the climatological average WBGT, creating twenty new WBGT values and recreating the daily maximum, mean, and minimum differences. This is repeated for each day and model (Fig. 1 ). Step #4: Selecting the best GCMs for each location We select a subset of six GCMs per location, ensuring GCMs with the best local skill are used in subsequent analysis. We rank GCMs by comparing each model’s WBGT cumulative distribution function (CDF) from 15 years of historical runs (2000–2014) to the WBGT CDF from the 14th WX data sets over the same period. Given our focus on high WBGT, comparing CDF tails (specifically the right “hot” tail) is more relevant than overall CDF shape in ranking GCMs. We use the two-sample Anderson-Darling (AD) test (Anderson and Darling 1952 ) implemented with R package “twosamples” (Dowd 2023 ). This test considers difference in CDFs shape and symmetry and is more sensitive to differences at distribution tails than other tests (Engmann and Cousineau 2011 ). We also compare WBGT provided by the 14th WX and WBGT calculated from weather observations using the Liljegren method (Fig. S8). Test statistics comparing the 14th WX data set with these WBGTs are much smaller than the test statistics between the 14th WX data set and any GCM CDF, providing additional evidence that the Liljegren method is robust for calculating WBGT since distributions closely align. Supplemental table S3 provides AD test statistics for each GCM output. Step #5: Empirical Quantile Mapping to correct GCM bias Bias adjustment is applied to account for systematic bias in climate models, particularly in regional and extreme events studies (Jeon, Paciorek et al. 2016 , Qian and Chang 2021 , Lehner, Nadeem et al. 2023 ). We use quantile delta mapping (QDM) implemented in R package “qmap” (Gudmundsson 2016 ) because QDM outperforms simpler methods when adjusting distribution tails (Qian and Chang 2021 , Lehner, Nadeem et al. 2023 ) and has been used in studies of climate change related health outcomes (Jeon, Paciorek et al. 2016 ). Our QDM implementation generates transformation functions unique to each GCM, scenario, and location between the observed and modelled WBGT CDFs (i.e., 14th WX data and GCM runs between 2015–2021) at every 0.5th quantile. These transformation functions are then applied to projected WBGT derived from GCMs (2025–2100). The 0.5 percentile step recognizes the small percentage of the CDF containing impactful (i.e., extreme) WBGT values and adjusts the CDF appropriately (Fig. S9). For example, at our coolest study location (FLW), significantly impactful WBGT > 31.11 o C occur < 1% of the time. A coarser transformation step would not appropriately adjust such a small percentage of values. Adjustments are made to each year’s CDF independently, retaining trends between sequential years. For example, the transformation is applied to WBGT values for 2025 to create a bias-adjusted 2025 data set, then reapplied to 2026 data and so on for subsequent years. This assumes CDFs for future year remain similarly shaped to the CDF derived from the 14th WX data but allows each year’s CDF to progressively shift warmer. Data sets are complete after bias adjustment. At each location our final data sets include scenario specific, bias corrected hourly WBGT from six GCMs for the periods of 2015–2021 and 2025–2100. Heat Wave and Trend Evaluation Our heat wave assessment criteria is a modification of the Russo et al ( 2014 ) Heat Wave Magnitude Index (HWMI). Inclusion criteria is 6 or more hours per day ≥31.11 o C WBGT for three or more consecutive days, reflecting our desire to evaluate heat waves for their impacts on outdoor labor. For nearly all outdoor labor and recreation, activity modification is recommended at or before 31.11 o C (United States. Occupational and Health 2017, Roberts, Armstrong et al. 2021 , HQDA 2022 ), and 6 hours of “lost” time (a quarter of the day) reflects the potential loss of a workday. We do not determine subheat wave magnitude but in the interest in assessing labor impacts, heat wave magnitude is the product of mean hours per day ≥ 31.11 o C and the length of the heat wave in days; the resulting HWMI approximates the cumulative hours ≥ 31.11 o C for each heat wave. Following Russo et al ( 2014 ), we choose HWMI as the maximum heat wave in a year. For projections, this is the greatest HWMI from any of the six GCMs. All trends are calculated using a seasonal variation of Mann-Kendall analysis, a robust method testing for monotonic trends in time series data (Donald Meals 2011 ). We implement this test using R package ‘EnvStats’ (Steven P. Millard 2022 ). To test trends in future weather variables we use an unadjusted multi-model mean of daily mean values from all 19 GCMs. RESULTS Evaluating calculated WBGT Our first evaluation compares WBGT calculated using the Liljegren method with values provided by the 14th WX. Since both are derived from the same observations the two data sets should closely match. We find distributions agree at all locations (Fig. 2 ) and same-hour WBGT comparison have r 2 values between 0.987–0.992 (Fig. S10). The next evaluation, correlating weather variable inputs and the 14th WX WBGT and Liljegren calculated WBGT values, find correlations nearly identical (Fig. S11). Comparing un-adjusted multi-model means using GCM historical runs between 2000–2014 and the 14th WX data returns differences between − 1.29 to 0.87 o C and multi-model median differences between − 2.20 to 0.88 o C WBGT (Table S12). Without selecting GCMs subset based on local skill and before bias adjusting, estimating WBGT from GCM output appears accurate; mean and median differences are reduced to < < 1 o C when comparing the selected and bias adjusted GCM subset outputs (2015–2021) for either scenario (Table S13). Correlations between input variables and output WBGT Considering the variables that influence WBGT provides insights to the local timing and conditions when high WBGT occurs. T a and specific humidity always correlate strongly with WBGT at all locations (Table S14). Daily mean T a , not maximum, is most strongly correlated to maximum WBGT at all locations while daily minimum T a is always correlated most strongly with daily minimum WBGT. At southeast locations (FMGA and FJSC) humidity is the variable second most strongly correlated with maximum WBGT, while at FSOK humidity is third (behind minimum T a ) and at FLW it is fourth (behind maximum and minimum T a ). Wind speed and solar radiation are weakly to moderately correlated with WBGT. Wind speed is always negatively correlated and is weakest at FSOK, the location with highest average velocities. Daily maximum and mean solar radiation are always positively correlated with WBGT, with correlation stronger with daily maximum WBGT than daily minimum, expected since minimum WBGT occurs prior to sunrise. Visualizing intra-year and inter-year trends show how weather variables may influence future WBGT (Fig. 3 ; quantified in Figs. S15a-S15d). Specific humidity and T a are projected to increase, driving a commensurate increase to the WBGT, while wind speed trends are small. There are no inter-year trends to solar radiation, but the timing of peak daily mean solar radiation is interesting in that it only occurs near the summer solstice at FLW; at the Southeast locations it occurs around mid-May. Seasonally and related factors such as cloud cover, humidity, and aerosols affect inter-year amounts of solar radiation incident on the surface. Intra-year peak solar radiation timing is consistent with estimates from solar power planning calculators at nearby locations (SEL 2024 ). Intra-year weather variable trends differ among locations, and understanding the synergistic timing of these trends provides insight to when, and for how long, WBGT may be elevated at a location. At FSOK, specific humidity peaks between June and July, while mean T a doesn’t peak until late July. The relative early increase and slow decline in humidity along with the later peak in T a results in an extended period of locally elevated WBGT compared to other locations, with mean WBGT peaking around the first week of July and remaining generally flat compared to curves at other installations (bottom row, Fig. 3 ). In contrast, mean WBGT peaks between 23–29 July (FJSC) and 26–31 July (FMGA) with specific humidity values peaking a few days prior and mean T a about a week earlier. At these locations WBGT peaks shortly after temperature and humidity because wind speed decreases through mid-August. At our fourth location (FLW), specific humidity, T a , and WBGT all peak near the same period (17–22 July) resulting in a narrower window of elevated mean WBGT. Future WBGT Distributions WBGT distribution displays surprising changes at some locations, particularly in the warmer SSP 3 scenario where a bimodal peak emerges at FJSC and FMGA in later years (Fig. 4 ). This peak is less well defined at FSOK and does not appear at FLW. Given influence of specific humidity and T a on WBGT, emerging peaks at the warm end of the distribution are likely related to interaction between these two variables. Dividing WBGT distribution by day and night demonstrates the bi-modal peaks originate during daylight hour (Fig. S16). The distribution of nighttime WBGT, less influenced by humidity trends since relative humidity ≅ 100% is more common at night, generally retain their shape while shifting warmer. WBGT index trends Daily minimum WBGT is projected to increase faster than daily mean WBGT, which increase faster than daily maximum WBGT; nights are warming faster than days (Table 2). This is consistent with trend assessments of T a within the USA (Vanos, Kalkstein et al. 2015 , USGCRP 2023 ). Table 2: Trends as the multi-model mean for daily maximum, mean, and minimum WBGT per decade. Consistent with scenario design (O’Neill, Kriegler et al. 2017 ), SSP 3 shows a continued rate of WBGT warming as the century progresses, while SSP 2 shows a positive but decreasing rate between the early and late periods. During the warm season (May-October) the rate of daily minimum WBGT increase is as high as 0.625 o C decade − 1 in SSP 3 at FSOK and 0.465 o C decade − 1 in SSP 2 at FLW. In contrast, the rate of increase for daily maximum WBGT is 0.451 o C decade − 1 in SSP 3 at FSOK and 0.323 o C decade − 1 in SSP 2 at FLW. DISCUSSION Increase in WBGT extremes and heatwave frequency We use heat category 4 (≥ 31.11 o C) to identify WBGT significant enough to cause severe labor impacts. This value exceeds recommended limits for acclimatized workers performing “light” work (NIOSH 2016 ), the Occupational Safety Health Administration recommend significant rest periods when working above this threshold (United States. Occupational and Health 2017) and consideration should be given to rescheduling exertional outdoor military training (Office of the Surgeon OSG 2023 ). The American College of Sports Medicine recommends rescheduling athletic competitions well before this threshold and limiting training to less than 1 hour for low-risk individuals (Roberts, Armstrong et al. 2021 ). Our results show significant increases in the frequently of hours exceeding 31.11 o C (Fig. 5 ). By 2039, depending on location, hours ≥ 31.11 o C increase between 22–63% above a 2015–2021 baseline. By mid-century the number of hours per year above this threshold approximately doubles (88–151%), with variations emerging between scenarios. By 2099, hours ≥ 31.11 o C increase at least 143% in the SSP 2 scenario and up to 443% in SSP 3. By century’s end, at each location hours above this threshold occur at least 100% more frequently in SSP 3 than in SSP 2 compared to baseline (Table S17). Our HWMI is a proxy for hot hours on consecutive days and may be more relevant for outdoor labor since it approximates hours potentially lost to outdoor labor during a heatwave (Table S18). Increases in consecutive days primarily drives increasing HWMI. At FLW, the location with the lowest heat burden, heatwaves historically averaged a HWMI of 28 and a duration of ~ 4 days long (i.e., ~ 28 hours ≥ 31.11 o C over four consecutive days). Based on projections, by 2039 the HWMI increases to 39 with an average length of 5 days; by the 2050s, a HWMI of 45 with a length of 6 days. In the SSP 2 scenario, by 2099 heatwaves are on average more than 8 days long at FLW. At the other three locations, heatwave duration increases from about 4 days to, on average, 20 days by mid-century, driven by the increase in consecutive days ≥ 31.11 o C. Hours per day ≥ 31.11 o C also increase but never exceeds 11 hours per day in any projection, likely due to cooling at night. Weeks long heatwaves will create significant challenges for outdoor labor, and particularly for outdoor labor conducted on fixed schedules. For instance, U.S. military recruit training follows a 10-week program, construction often consists of sequential activities, and agricultural work depends on specific growing cycles and seasonal timing. These types of activities will struggle under extended periods of activity-limiting heat. The most extreme heatwave predicted in any model and scenario occurs at FMGA near the end of the century. At 64 days (over 9 weeks) long, during which 707 hours (most daytime hours) are ≥ 31.11 o C, such an event represents less of a heatwave and more of a permanent shift toward higher summer WBGT and would require extreme adaptations to any schedule dependent outdoor labor program. In some cases, labor can shift to cooler nighttime periods, but this presents a new series of challenges, including additional health risks (He, Kim et al. 2022 ). The lowest threshold in our study, WBGT ≥ 25.56 o C, still presents risk for EHI. Individuals predisposed to EHI and those from cooler climates are at increased risk during athletic training and competition even at this lower threshold (Roberts, Armstrong et al. 2021 ) while those performing heavy labor are recommended to take rest breaks that can exceed 50% of total labor time (United States. Occupational and Health 2017, HQDA 2022 ). The effect of temperature increases on workplace injury rates correlate more with overnight lows compared with daytime highs (McInnes, Akram et al. 2017 ). Historically few nights at our locations had minimum temperature ≥ 25.56 o C and GCM historical runs average 1 or fewer such nights per year. Such infrequent occurrence won’t remain the norm. Daily minimum WBGT are projected to increase across all locations, especially FSOK. In the SSP 2 scenario, three locations project ~ 5 nights per year with minimum WBGT ≥ 25.56 o C between 2050–2100. In contrast, FSOK projections frequently indicate 20 + days per year in the last two decades of the century. Study Limitations WBGT index is sensitive to microclimates (Clark and Konrad 2024 ). Solar, wind, or humidity patterns may differ between nearby locations depending on local topography. Our observations come from a single weather station. Projected WBGT is thus qualified as accurate for these point locations, although trends in surrounding areas are likely similar. Using climatological average values (reference step #1) introduces uncertainty. The 95th CI for hourly mean values is generally small during warm months (< 2 o C), but this uncertainty is carried forward to projected WBGT (Table S19). Our bias adjustment method assumes future WBGT distributions can be approximated from past distributions. Despite limitations, we believe this study provides a precise methodology for projecting WBGT at point locations from GCM data due to our explicit calculations of WBGT and use of long-term local averages. Two further challenges limit implement at large spatial scales. Large computational effort is needed; here, we propose no solutions, since the Liljegren method requires two iterative calculations for each WBGT value. The second is the requirement for historical observations. With minimal modification, this method can theoretically be used in the absence of historical weather station records by replacing records with reanalysis data to calculate historic WBGT. Local climatological average hourly WBGT could thus be estimated (reference step #1), allowing this method to be applied in the absence of historical WBGT data sets. CONCLUSION Our study provides a method to project WBGT from GCM outputs. We test a WBGT estimation method developed by Liljegren by comparing outputs with WBGT values provided by the 14th WX, then by using the method with historical GCM run outputs instead of weather observations. WBGT is projected using GCM outputs from the NASA NEX-GDDP project by calculating future WBGT at sub-daily resolution and adjusting historical climatological average diurnal cycles to match these values, thereby projecting future daily WBGT diurnal curves. We select GCMs and conduct bias adjustment against historical observations to ensure outputs are derived from GCMs with the best local skill and reflect the correct distribution of warm extreme values. The result at each location is hourly WBGT between 01/01/2025–12/31/2100 for two SSP scenarios from 6 GCMs. We quantify trends to WBGT and weather variables that influence WBGT, finding WBGT overnight lows increasing fastest at all locations. Weather variables trend similarly across all locations with the timing of intra-year maximum and minimum unique to each location. We consider safety thresholds designed to mitigate EHI, finding the frequency of daytime hours above 31.11 o C, the duration and intensity of heat waves, and the number of nights above 25.56 o C projected to increase. By quantifying these trends at high spatiotemporal resolution, we hope to provide a methodology to inform locally relevant adaptations required to continue outdoor labor and training despite climate change. Declarations The authors have no competing interests to declare that are relevant to the content of this article. Funding Acknowledgements Patton is funded by the Advanced Strategic Planning and Policy Program, Command and General Staff College, U.S. Army. Data Availability and Transparency Data sets and code can be downloaded at: https://doi.org/10.7924/r4st7st5t Author Contributions All authors contributed to the study. Methodology design was performed by Patton and Li. Data generation and initial analysis was performed by Patton. The first draft of the manuscript was written by Patton and all authors were involved in manuscript revisions. All authors read and approved the final manuscript. References Ahn, Y., C. K. Uejio, J. Rennie and L. Schmit (2022). "Verifying Experimental Wet Bulb Globe Temperature Hindcasts Across the United States." GeoHealth 6 (4): e2021GH000527. Alele, F. O., B. S. Malau-Aduli, A. E. O. Malau-Aduli and J. C. M (2020). "Epidemiology of Exertional Heat Illness in the Military: A Systematic Review of Observational Studies." Int J Environ Res Public Health 17 (19). Anderson, T. W. and D. A. Darling (1952). "Asymptotic Theory of Certain "Goodness of Fit" Criteria Based on Stochastic Processes." The Annals of Mathematical Statistics 23 (2): 193-212. Blazejczyk, K., Y. Epstein, G. Jendritzky, H. Staiger and B. Tinz (2012). "Comparison of UTCI to selected thermal indices." International Journal of Biometeorology 56 (3): 515-535. Brimicombe, C., C. H. B. Lo, F. Pappenberger, C. Di Napoli, P. Maciel, T. Quintino, R. Cornforth and H. L. Cloke (2023). "Wet Bulb Globe Temperature: Indicating Extreme Heat Risk on a Global Grid." GeoHealth 7 (2): e2022GH000701. Budd, G. M. (2008). "Wet-bulb globe temperature (WBGT)--its history and its limitations." J Sci Med Sport 11 (1): 20-32. Buzan, J. R. and M. Huber (2020). "Moist Heat Stress on a Hotter Earth." Annual Review of Earth and Planetary Sciences 48 (1): 623-655. Cai, J. (2019). humidity: Calculate Water Vapor Measures from Temperature and Dew Point. CRAN. 0.1.5 . Chavaillaz, Y., P. Roy, A.-I. Partanen, L. Da Silva, É. Bresson, N. Mengis, D. Chaumont and H. D. Matthews (2019). "Exposure to excessive heat and impacts on labour productivity linked to cumulative CO2 emissions." Scientific Reports 9 (1): 13711. Clark, J. and C. E. Konrad (2024). "Observations and Estimates of Wet-Bulb Globe Temperature in Varied Microclimates." Journal of Applied Meteorology and Climatology 63 (2): 305-319. Donald Meals, J. S., Steven Dressing, Jon Harcum (2011). Statistical Analysis for Monotonic Trends. Tech Notes , National Nonpoint Source Monitoring Program (NNPSMP). Dowd, C. (2023). twosample. CRAN. Engmann, S. and D. Cousineau (2011). "Comparing distributions: the two-sample Anderson–Darling test as an alternative to the Kolmogorov–Smirnov test." Journal of Applied Quantitative Methods 6 : 1-17. Gudmundsson, L. (2016). Statistical Transformations for Post-Processing Climate Model Output : Empirical adjustment of the distribution of variables originating from (regional) climate model simulations using quantile mapping. Hausfather, Z. (2018). "Explainer: How ‘Shared Socioeconomic Pathways’ explore future climate change." Retrieved February 2nd,, 2024, from https://www.carbonbrief.org/explainer-how-shared-socioeconomic-pathways-explore-future-climate-change/. Hausfather, Z. and G. P. Peters (2020). "Emissions - the 'business as usual' story is misleading." Nature 577 (7792): 618-620. He, C., H. Kim, M. Hashizume, W. Lee, Y. Honda, S. E. Kim, P. L. Kinney, A. Schneider, Y. Zhang, Y. Zhu, L. Zhou, R. Chen and H. Kan (2022). "The effects of night-time warming on mortality burden under future climate change scenarios: a modelling study." The Lancet Planetary Health 6 (8): e648-e657. Hosokawa, Y., D. J. Casa, J. M. Trtanj, L. N. Belval, P. A. Deuster, S. M. Giltz, A. J. Grundstein, M. D. Hawkins, R. A. Huggins, B. Jacklitsch, J. F. Jardine, H. Jones, J. B. Kazman, M. E. Reynolds, R. L. Stearns, J. K. Vanos, A. L. Williams and W. J. Williams (2019). "Activity modification in heat: critical assessment of guidelines across athletic, occupational, and military settings in the USA." International journal of biometeorology 63 (3): 405-427. HQDA (2021). AR 115-10/AFI 15-157 Weather Support for the U.S. Army. Washington DC, Department of the Army and Department of the Air Force. HQDA (2022). TB MED 507: Heat Stress Control and Heat Casualty Management. Washington DC, Headquarters, Department of the Army. IPCC (2021). Sixth Assessment Report Working Group II - Impacts, Adaptation and Vulnerability. Sixth Assessment Report , Intergovernmental Panel on Climate Change. IPCC (2023). Climate Change 2023: Synthesis Report. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. J. R. H. Lee. Geneva, Switzerland. Jeon, S., C. J. Paciorek and M. F. Wehner (2016). "Quantile-based bias correction and uncertainty quantification of extreme event attribution statements." Weather and Climate Extremes 12 : 24-32. Knutson, T. R. and J. J. Ploshay (2016). "Detection of anthropogenic influence on a summertime heat stress index." Climatic Change 138 (1): 25-39. Kong, Q. and M. Huber (2022). "Explicit Calculations of Wet-Bulb Globe Temperature Compared With Approximations and Why It Matters for Labor Productivity." Earth's Future 10 (3): e2021EF002334. Kriegler, E., N. Bauer, A. Popp, F. Humpenöder, M. Leimbach, J. Strefler, L. Baumstark, B. L. Bodirsky, J. Hilaire, D. Klein, I. Mouratiadou, I. Weindl, C. Bertram, J.-P. Dietrich, G. Luderer, M. Pehl, R. Pietzcker, F. Piontek, H. Lotze-Campen, A. Biewald, M. Bonsch, A. Giannousakis, U. Kreidenweis, C. Müller, S. Rolinski, A. Schultes, J. Schwanitz, M. Stevanovic, K. Calvin, J. Emmerling, S. Fujimori and O. Edenhofer (2017). "Fossil-fueled development (SSP5): An energy and resource intensive scenario for the 21st century." Global Environmental Change 42 : 297-315. Lehner, F., I. Nadeem and H. Formayer (2023). "Evaluating skills and issues of quantile-based bias adjustment for climate change scenarios." Adv. Stat. Clim. Meteorol. Oceanogr. 9 (1): 29-44. Lemke, B. and T. Kjellstrom (2012). "Calculating Workplace WBGT from Meteorological Data: A Tool for Climate Change Assessment." Industrial Health 50 (4): 267-278. Lewandowski, S. A., M.-A. Kioumourtzoglou and J. L. Shaman (2022). "Heat stress illness outcomes and annual indices of outdoor heat at U.S. Army installations." PLOS ONE 17 (11): e0263803. Li, C., X. Zhang, F. Zwiers, Y. Fang and A. M. Michalak (2017). "Recent Very Hot Summers in Northern Hemispheric Land Areas Measured by Wet Bulb Globe Temperature Will Be the Norm Within 20 Years." Earth's Future 5 (12): 1203-1216. Lieblich, M. (2017). mdljts/wbgt. GitHub, University of Washington. Liljegren, J. C., R. A. Carhart, P. Lawday, S. Tschopp and R. Sharp (2008). "Modeling the wet bulb globe temperature using standard meteorological measurements." J Occup Environ Hyg 5 (10): 645-655. McInnes, J. A., M. Akram, E. M. MacFarlane, T. Keegel, M. R. Sim and P. Smith (2017). "Association between high ambient temperature and acute work-related injury: a case-crossover analysis using workers' compensation claims data." Scand J Work Environ Health 43 (1): 86-94. Minard, D. (1961). "Prevention of Heat Casualties in Marine Corps Recruits: Period of 1955–60, With Comparative Incidence Rates And Climatic Heat Stresses in Other Training Categories." Military Medicine 126 (4): 261-272. NIOSH (2016). Occupational Exposure to Heat and Hot Environments Revised Criteria 2016. D. o. H. a. H. Services. Concinnati, OH, Centers for Disease Control and Prevention, National Insititude for Occupational Safety and Health. NMFS (2023). NATIONAL MARINE FISHERIES SERVICE PROCEDURE 02-110-18. DoC. USA, National Marine Fisheries Service, Department of Commerce. O’Neill, B. C., E. Kriegler, K. L. Ebi, E. Kemp-Benedict, K. Riahi, D. S. Rothman, B. J. van Ruijven, D. P. van Vuuren, J. Birkmann, K. Kok, M. Levy and W. Solecki (2017). "The roads ahead: Narratives for shared socioeconomic pathways describing world futures in the 21st century." Global Environmental Change 42 : 169-180. OSG (2023). Heat Illness Prevention for the 2023-2024 heat season. Washington, DC, Office of the Surgeon General, Department of the Army. Parsons, K. (2006). "Heat stress standard ISO 7243 and its global application." Ind Health 44 (3): 368-379. Parsons, L. A., D. Shindell, M. Tigchelaar, Y. Zhang and J. T. Spector (2021). "Increased labor losses and decreased adaptation potential in a warmer world." Nat Commun 12 (1): 7286. Patel, T., S. P. Mullen and W. R. Santee (2013). "Comparison of Methods for Estimating Wet-Bulb Globe Temperature Index From Standard Meteorological Measurements." Military Medicine 178 (8): 926-933. Patton, E. M. and M. W. Doyle (2023). "Observed Warming Trends at U.S. Army Basic Combat Training Installations and Implications for Future Recruit Training." Mil Med . Périard, J. D., D. DeGroot and O. Jay (2022). "Exertional heat stroke in sport and the military: epidemiology and mitigation." Exp Physiol 107 (10): 1111-1121. Pisaniello, D. and R. Di Corleto (2023). Chapter 6 - Heat exposure and occupational health. Heat Exposure and Human Health in the Context of Climate Change . Y. Guo and S. Li, Elsevier : 109-131. Qian, W. and H. H. Chang (2021). "Projecting Health Impacts of Future Temperature: A Comparison of Quantile-Mapping Bias-Correction Methods." Int J Environ Res Public Health 18 (4). Riahi, K., D. P. van Vuuren, E. Kriegler, J. Edmonds, B. C. O’Neill, S. Fujimori, N. Bauer, K. Calvin, R. Dellink, O. Fricko, W. Lutz, A. Popp, J. C. Cuaresma, S. Kc, M. Leimbach, L. Jiang, T. Kram, S. Rao, J. Emmerling, K. Ebi, T. Hasegawa, P. Havlik, F. Humpenöder, L. A. Da Silva, S. Smith, E. Stehfest, V. Bosetti, J. Eom, D. Gernaat, T. Masui, J. Rogelj, J. Strefler, L. Drouet, V. Krey, G. Luderer, M. Harmsen, K. Takahashi, L. Baumstark, J. C. Doelman, M. Kainuma, Z. Klimont, G. Marangoni, H. Lotze-Campen, M. Obersteiner, A. Tabeau and M. Tavoni (2017). "The Shared Socioeconomic Pathways and their energy, land use, and greenhouse gas emissions implications: An overview." Global Environmental Change 42 : 153-168. Roberts, W. O., L. E. Armstrong, M. N. Sawka, S. W. Yeargin, Y. Heled and F. G. O'Connor (2021). "ACSM Expert Consensus Statement on Exertional Heat Illness: Recognition, Management, and Return to Activity." Curr Sports Med Rep 20 (9): 470-484. Russo, S., A. Dosio, R. G. Graversen, J. Sillmann, H. Carrao, M. B. Dunbar, A. Singleton, P. Montagna, P. Barbola and J. V. Vogt (2014). "Magnitude of extreme heat waves in present climate and their projection in a warming world." Journal of Geophysical Research Atmospheres 119 (22): 12,500-512,512. Sawka, M. N., L. R. Leon, S. J. Montain and L. A. Sonna (2011). "Integrated physiological mechanisms of exercise performance, adaptation, and maladaptation to heat stress." Compr Physiol 1 (4): 1883-1928. Scafetta, N. (2024). "Impacts and risks of “realistic” global warming projections for the 21st century." Geoscience Frontiers 15 (2): 101774. SEL. (2024). "Solar Energy & Solar Power Resources." Retrieved February 16th, 2024, from https://www.solarenergylocal.com. Spangler, K. R., S. Liang and G. A. Wellenius (2022). "Wet-Bulb Globe Temperature, Universal Thermal Climate Index, and Other Heat Metrics for US Counties, 2000–2020." Scientific Data 9 (1): 326. Steven P. Millard, A. K. (2022). EnvStats: Package for Environmental Statistics, Including US EPA Guidance. Takakura, J., S. Fujimori, K. Takahashi, Y. Hijioka and Y. Honda (2019). "Site-specific hourly resolution wet bulb globe temperature reconstruction from gridded daily resolution climate variables for planning climate change adaptation measures." Int J Biometeorol 63 (6): 787-800. Thrasher, B. (2023). NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP-CMIP6) Technical Note, NASA. Thrasher, B., Wang, W., Michaelis, A. Nemani, R. (2021). NEX-GDDP-CMIP6. NASA. Center for Climate Simulation. United States. Occupational, S. and A. Health (2017). OSHA technical manual . Washington, D.C., U.S. Dept. of Labor, Occupational Safety and Health Administration. USGCRP (2023). Fifth National Climate Assessment . Washington, DC, USA. Vanos, J. K., L. S. Kalkstein and T. J. Sanford (2015). "Detecting synoptic warming trends across the US Midwest and implications to human health and heat-related mortality." International Journal of Climatology 35 (1): 85-96. Willett, K. M. and S. Sherwood (2012). "Exceedance of heat index thresholds for 15 regions under a warming climate using the wet-bulb globe temperature." International Journal of Climatology 32 (2): 161-177. Yeargin, S., R. Hirschhorn, A. Grundstein, D. Arango, A. Graham, A. Krebs and S. Turner (2023). "Variations of wet-bulb globe temperature across high school athletics in South Carolina." International Journal of Biometeorology . Cite Share Download PDF Status: Published Journal Publication published 17 Sep, 2024 Read the published version in International Journal of Biometeorology → Version 1 posted Reviewers agreed at journal 22 May, 2024 Reviewers invited by journal 22 May, 2024 Editor assigned by journal 16 May, 2024 First submitted to journal 16 May, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4414813","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":305603150,"identity":"8044ae0b-faf8-4377-9ae6-82995bb2413e","order_by":0,"name":"Erik Patton","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBElEQVRIiWNgGAWjYFACxmYkjoGEHIg68IAoLWzMQKLCwhisJQG/NcxIWs5UJDaAOPi08EsfbjZgzGHI45/ff/BxYZtE+vywww+BttjJ6TZg1yLZl9icwLiNoVjiGDOz8cw2idyNt9MMgFqSjc0OYNdicIax+QBQS2LDMWY2aV6QltkJIC0HErfh0GIP0zIfqiXdcHb6B7xaDHgYwQ5L3ADSwnNGIkFeOge/LRJAWwwSt0kUGx5LNjbmqZAw3CCdU3AgwQC3X/h72B9LfNxmkyd3+ODDxzwGdfLys9M3f/hQYSeHSwsYJDBIJCCcClZpgEc5XBcMyDcQVj0KRsEoGAUjCwAAx99Zwus+l50AAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-0628-7706","institution":"Duke University Nicholas School of the Environment","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Erik","middleName":"","lastName":"Patton","suffix":""},{"id":305603151,"identity":"75e30474-c062-4954-bd9d-4be430972e61","order_by":1,"name":"Wenhong Li","email":"","orcid":"","institution":"","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Wenhong","middleName":"","lastName":"Li","suffix":""},{"id":305603152,"identity":"51a4788c-7bc2-424a-aa58-38031408678d","order_by":2,"name":"Ashley Ward","email":"","orcid":"","institution":"","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Ashley","middleName":"","lastName":"Ward","suffix":""},{"id":305603153,"identity":"ce71abbe-9a52-4cba-8fcc-ac472bfc2eae","order_by":3,"name":"Martin Doyle","email":"","orcid":"","institution":"","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Martin","middleName":"","lastName":"Doyle","suffix":""}],"badges":[],"createdAt":"2024-05-13 18:04:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4414813/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4414813/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00484-024-02776-5","type":"published","date":"2024-09-17T15:57:20+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":57693759,"identity":"fcc5259b-8d6b-4524-a0e7-8a7a4a21cff4","added_by":"auto","created_at":"2024-06-04 12:01:47","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":866540,"visible":true,"origin":"","legend":"\u003cp\u003eCreating daily diurnal curves from GCM maximum, mean, and minimum WBGTs\u003cstrong\u003e.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eDiamonds = daily maximum values. Triangles = daily mean values (up is morning, down is afternoon). Circles = daily minimum values. Blue symbols are climatological average, purple and green are two separate GCM outputs. X-axis is GMT time. For simplicity, only two GCMs are shown in the center two panels, connecting maximum, mean, and minimum WBGT index by dashed lines. Bottom panel shows all 19 GCMs. The overall shape of the daily diurnal cycle is well preserved in most models and on most days, although the hour of maximum (minimum) WBGT index occurrence may shift slightly earlier/later. Curves shown using data before bias adjustment.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-4414813/v1/38553cfd26274e5883891f85.png"},{"id":57693176,"identity":"cc57984b-6665-4659-a2ab-b8904f4104d7","added_by":"auto","created_at":"2024-06-04 11:53:46","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":273568,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation between WBGT calculated using Liljegren method and WBGT provided by the 14\u003csup\u003eth\u003c/sup\u003e Weather Squadron.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAF Method = WBGT provided by the U.S. Air Force 14\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e Weather Squadron. Locations from the top: Ft Jackson, South Carolina (FJSC); Ft Moore, Georgia (FMGA); Ft Sill, Oklahoma (FSOK); Ft Leonard Wood, Missouri (FLW). Left columns overlays the distributions of hourly WBGT calculated by the Liljegren method with hourly WBGT provided by the 14\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e Weather Squadron. Right column overlays the mean WBGT at each hour of the year (n=8,760) derived from hourly WBGT between 2000-2014 (n=131,400). In both columns, the two data sets overlap extensively.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-4414813/v1/721495978a7868df2b05773e.png"},{"id":57693175,"identity":"da4235f8-eff0-4795-9777-7b2477cebc60","added_by":"auto","created_at":"2024-06-04 11:53:46","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1066526,"visible":true,"origin":"","legend":"\u003cp\u003eDaily mean values for input variables and WBGT averaged across 19 GCMs, 2025-2100.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eColumns from left to right: Ft Jackson, SC; Ft Moore, GA; Ft Sill, OK; Ft Leonard Wood, MO. Variable change across the year is shown by the shape of the curve, starting January 1\u003c/em\u003e\u003csup\u003e\u003cem\u003est\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e (far left; day 1) and ending on December 31\u003c/em\u003e\u003csup\u003e\u003cem\u003est\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e (far right; day 365). Julian day 200 (x-axis) corresponds with July 19\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e. Projected inter-year changes are shown grading from 2025 (yellow) to 2100 (red). The dashed line is centered on August 4\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e, approximately the latest day that mean WBGT historically peaks during any year. Historically, the last week of July to the first week of August are the warmest periods on the WBGT index. Note y-axis scales differ between the first three rows in each column.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-4414813/v1/8a7d65ace747966674951cb9.png"},{"id":57693178,"identity":"8c108290-17a7-4650-b1a1-51965bfc6a64","added_by":"auto","created_at":"2024-06-04 11:53:47","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":242764,"visible":true,"origin":"","legend":"\u003cp\u003eWBGT index distributions at four periods.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eLeft column: WBGT distribution from 14\u003c/em\u003e\u003csup\u003e\u003cem\u003eth\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e WX (“observations”) and GCM SSP 2 and SSP 3 historical runs between 2015-2021. Distributions are virtually identical. Center column: SSP 2 projected WBGT index for decades centered on 2030, 2050, and 2090. Right Column: Same as center column but for SSP 3. Rows top to bottom: Ft Jackson, SC; Ft Moore, GA; Ft Sill, OK; Ft Leonard Wood, MO.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-4414813/v1/fa998209245ce3d4cf46463d.png"},{"id":57693177,"identity":"ae04b6ad-a5ff-4088-8b58-f91b3836d01a","added_by":"auto","created_at":"2024-06-04 11:53:46","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":960856,"visible":true,"origin":"","legend":"\u003cp\u003eTrends to relevant WBGT thresholds\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e(A) Trends to hours per year above 31.11\u003c/em\u003e\u003csup\u003e\u003cem\u003eo\u003c/em\u003e\u003c/sup\u003e\u003cem\u003eC (red flag conditions). Shaded uncertainty created using the models with greatest and least hours per year. (B) Trends to our modified Heat Wave Magnitude Index. (C) Trends in the number of days with overnight lows exceeding 25.56\u003c/em\u003e\u003csup\u003e\u003cem\u003eo\u003c/em\u003e\u003c/sup\u003e\u003cem\u003eC (white flag conditions). (C) panels are created from GCM scenarios (not observations) beginning in 2015. Our data sets exclude 2022-2024.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-4414813/v1/cb8b3e53334aff306df80fb2.png"},{"id":65103998,"identity":"6830f862-4ff4-4e4e-979a-bf0f7d5f8668","added_by":"auto","created_at":"2024-09-23 16:10:36","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3662392,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4414813/v1/f79b54c6-78c5-4b5b-aca7-dd8a61589421.pdf"}],"financialInterests":"","formattedTitle":"Wet bulb globe temperature from climate model outputs: a method for projecting hourly site-specific values and trends.","fulltext":[{"header":"Introduction","content":"\u003cp\u003eClimate change and associated increasing global temperatures (IPCC \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) will affect many aspects of society, including outdoor labor. However, local impacts of climate change and warming depend on regional geographies, weather patterns, and local conditions. Understanding the local impact of climate change on outdoor labor thus demands local scale analysis. Additionally, temperature is rarely the best index to measure thermal influences on human health (Li, Zhang et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2017\u003c/span\u003e, Buzan and Huber \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2020\u003c/span\u003e, Pisaniello and Di Corleto 2023). Factors such as humidity, wind, sun exposure, and exertional effort are relevant for outdoor workers. To consider these variables on thermal stress, over 100 variants of temperature indices have been developed (Blazejczyk, Epstein et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Of these, the wet bulb globe temperature (WBGT) index is widely used to establish activity modification thresholds to prevent heat illness (Budd \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Kong and Huber \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). For example, WBGT is used to regulate athletic practice and competition (Roberts, Armstrong et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Yeargin, Hirschhorn et al. \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), establish occupational health and safety standards (Parsons \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2006\u003c/span\u003e, P\u0026eacute;riard, DeGroot et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and modify training in military settings (Minard \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e1961\u003c/span\u003e, HQDA \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Over most of the northern hemisphere, new high summer mean WBGT records are projected to occur more frequently than new mean high records measured by surface temperature alone (Li, Zhang et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWBGT is the weighted average of the natural wet bulb temperature (\u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ewb\u003c/em\u003e\u003c/sub\u003e), dry bulb temperature (\u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e), and black globe temperature (\u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ebg\u003c/em\u003e\u003c/sub\u003e) (Eq.\u0026nbsp;1). Each of these WBGT variables respond to weather variables differently. \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e is measured by a shaded thermometer. \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ewb\u003c/em\u003e\u003c/sub\u003e cools by latent heat loss and is a proxy for sweating, which removes heat from the skin during sweat evaporation. Latent heat loss is influenced by humidity, with less evaporation (and less heat removed) during periods of high relative humidity. \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ebg\u003c/em\u003e\u003c/sub\u003e is influenced by radiant heat sources, primarily solar radiation, and is measured inside a 6-inch black metal sphere (Budd \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2008\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eEquation 1: Wet Bulb Globe Temperature\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$WBGT Index=0.7\\left({T}_{wb}\\right)+0.2\\left({T}_{bg}\\right)+0.1\\left({T}_{a}\\right)$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThe WBGT index is often used to establish safety and activity modification thresholds for populations at risk of exertional heat illness (Hosokawa, Casa et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). During physical labor skeletal muscles generate large heat loads (Sawka, Leon et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) which, if not dissipated, increase the body\u0026rsquo;s core temperature. This can lead to exertional heat illness (EHI) ranging from heat cramps to potentially fatal heat stroke (Alele, Malau-Aduli et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). When ambient temperature exceeds skin temperature of ~\u0026thinsp;35\u003csup\u003eo\u003c/sup\u003eC, sweat evaporation becomes the only heat dissipation method possible (Sawka, Leon et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Heat loss from sweat evaporation decreases with increasing relative humidity until no evaporation, and therefore no net heat loss, which occurs at relative humidity of 100%. Heat stress is thus maximized when temperature and humidity are high and is additionally compounded by radiant solar heat when outdoors.\u003c/p\u003e \u003cp\u003eSafety thresholds use WBGT to protect against EHI are often calibrated for specific populations, such as the guidelines set forth by the American College of Sports Medicine (ACSM) for athletes, worker safety recommendations from the Occupational Safety and Health Administration (OSHA), and work-rest cycles for military personnel set by the U.S. Department of the Army. The ACSM recommends activity modification when WBGT exceeds values as low as 15.1\u003csup\u003eo\u003c/sup\u003eC and competition cancellation when WBGT exceeds values between 24.6\u0026ndash;32.3\u003csup\u003eo\u003c/sup\u003eC (Roberts, Armstrong et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) while the U.S. military modifies some outdoor training beginning at a WBGT of 25.56\u003csup\u003eo\u003c/sup\u003eC (HQDA \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). This study uses values from the military scale when assessing impacts to outdoor labor and training (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;line-height:150%;'\u003e\u003cstrong\u003eTable 1.\u003c/strong\u003e WBGT index thresholds used to guide activity modification by the U.S. Military.\u003c/p\u003e\n\u003ctable style=\"border: none;width:426.0pt;border-collapse:collapse;\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 87pt;border: 1pt solid windowtext;padding: 0in 5.4pt;height: 16pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e\u0026nbsp;\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 70.25pt;border-top: 1pt solid windowtext;border-right: 1pt solid windowtext;border-bottom: 1pt solid windowtext;border-image: initial;border-left: none;padding: 0in 5.4pt;height: 16pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e\u0026nbsp;\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 124.75pt;border-top: 1pt solid windowtext;border-right: 1pt solid windowtext;border-bottom: 1pt solid windowtext;border-image: initial;border-left: none;padding: 0in 5.4pt;height: 16pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003eWBGT index, \u003csup\u003eo\u003c/sup\u003eF\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 2in;border-top: 1pt solid windowtext;border-right: 1pt solid windowtext;border-bottom: 1pt solid windowtext;border-image: initial;border-left: none;padding: 0in 5.4pt;height: 16pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003eWBGT index, \u003csup\u003eo\u003c/sup\u003eC\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 87pt;border-right: 1pt solid windowtext;border-bottom: 1pt solid windowtext;border-left: 1pt solid windowtext;border-image: initial;border-top: none;padding: 0in 5.4pt;height: 16pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003eHeat Category\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 70.25pt;border-top: none;border-left: none;border-bottom: 1pt solid windowtext;border-right: 1pt solid windowtext;padding: 0in 5.4pt;height: 16pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003eFlag Color\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:58.5pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003eminimum\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:66.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003emaximum\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.65pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003eminimum\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:73.35pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003emaximum\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width:87.0pt;border:solid windowtext 1.0pt;border-top:none;padding: 0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e1\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003eWhite\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:58.5pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e78\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:66.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e81.9\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.65pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e25.56\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:73.35pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e27.77\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width:87.0pt;border:solid windowtext 1.0pt;border-top:none;background: #92D050;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e2\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:#92D050;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003eGreen\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:58.5pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:#92D050;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e82\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:66.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:#92D050;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e84.9\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.65pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:#92D050;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e27.78\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:73.35pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:#92D050;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e29.43\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width:87.0pt;border:solid windowtext 1.0pt;border-top:none;background: yellow;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e3\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:yellow;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003eYellow\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:58.5pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:yellow;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e85\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:66.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:yellow;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e87.9\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.65pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:yellow;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e29.44\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:73.35pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:yellow;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e31.10\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width:87.0pt;border:solid windowtext 1.0pt;border-top:none;background: red;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e4\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:red;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003eRed\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:58.5pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:red;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e88\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:66.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:red;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e89.9\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.65pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:red;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e31.11\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:73.35pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:red;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:black;'\u003e32.21\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width:87.0pt;border:solid windowtext 1.0pt;border-top:none;background: black;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cstrong\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:white;'\u003e5\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:black;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:white;'\u003eBlack\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:58.5pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:black;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:white;'\u003e90\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:66.25pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:black;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:white;'\u003e-\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:70.65pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:black;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:white;'\u003e32.22\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:73.35pt;border-top:none;border-left:none;border-bottom:solid windowtext 1.0pt;border-right:solid windowtext 1.0pt;background:black;padding:0in 5.4pt 0in 5.4pt;height:16.0pt;\"\u003e\n \u003cp style='margin:0in;font-size:16px;font-family:\"Calibri\",sans-serif;text-align:center;'\u003e\u003cspan style='font-family:\"Times New Roman\",serif;color:white;'\u003e-\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\u003cp\u003eWBGT can be accurately calculated from past weather observations (Liljegren, Carhart et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, Lemke and Kjellstrom \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2012\u003c/span\u003e, Patel, Mullen et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2013\u003c/span\u003e, Kong and Huber \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), but limitations in the spatiotemporal resolution of global circulation models (GCMs) and the large computational effort needed to explicitly calculate WBGT from meteorological observations in large data sets (Brimicombe, Lo et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) complicate projecting future WBGT accurately. Explicitly calculating WBGT requires five weather variables and iterative calculations to resolve both \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({T}_{bg}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({T}_{wb}\\)\u003c/span\u003e\u003c/span\u003e. GCMs typically provide such variables as daily mean values, but greater temporal resolution is needed to estimate local hourly impacts. In addition, WBGT is often sensitive to small scale local features, but GCM output is provided at coarse spatial resolution. Spatial resolution challenges can be partially mitigated by using downscaled GCM output, but the highest resolution data sets with all required variables is the 0.25 x 0.25 degree (approximately 25km by 25km) output from the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) (Thrasher 2021), which may still mask the impact of local conditions on projected WBGT.\u003c/p\u003e \u003cp\u003ePrevious studies commonly address these challenges by omitting the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({T}_{bg}\\)\u003c/span\u003e\u003c/span\u003e and considering regional or global trends (Willett and Sherwood \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2012\u003c/span\u003e, Knutson and Ploshay \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2016\u003c/span\u003e, Li, Zhang et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2017\u003c/span\u003e, Parsons, Shindell et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Omitting \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({T}_{bg}\\)\u003c/span\u003e\u003c/span\u003e is used to estimate indoor WBGT and simplify calculations but can underestimate outdoor WBGT. Methods that include \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({T}_{bg}\\)\u003c/span\u003e\u003c/span\u003e exist and have been applied to global gridded data sets (Brimicombe, Lo et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) yet may not capture local variations due to coarse spatial resolution, failing to adequately project WBGT at a specific location of interest. Takakura et al (\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) construct site-specific WBGT using a method conceptually similar to ours but simplify the computational effort by applying regressions to estimate WBGT from GCM outputs instead of explicitly calculating it; our method appears more accurate at recreating historical WBGT observations, and this additional accuracy is presumably carried forward when estimating future WBGT values.\u003c/p\u003e \u003cp\u003eThis study describes a novel method for providing site-specific WBGT projections based on historical observations and downscaled GCM output. Our key contribution is in removing many assumptions required in previous work by explicitly calculating WBGT, including the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({T}_{bg}\\)\u003c/span\u003e\u003c/span\u003e component, doing so at a local scale. We do this by evaluating the daily maximum, mean, and minimum WBGT on future days using downscaled GCM outputs. We make assumptions only for variables not included in the downscaled GCM data sets (e.g., surface pressure); these assumptions are constrained by historical observations. Projected WBGT values on intervening hours are then interpolated using location-specific historical climatological averages. From an initial large ensemble of GCMs, a subset of models selected for skill in recreating historical extreme value observations is then bias adjusted using historical observations to ensure a site-specific final output.\u003c/p\u003e \u003cp\u003eWe test out method against records from four U.S. army training installations: Ft Jackson, South Carolina (FJSC); Ft Moore, Georgia (FMGA); Ft Leonard Wood, Missouri (FLW); and Ft Sill, Oklahoma (FSOK) (Fig \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Choosing these locations allows us to compare historical WBGT records provided by the U.S Air Force 14th Weather Squadron (14th WX) with WBGT calculated from our methodology. We then apply our method using GCMs selected for skill at each location and bias-adjust the output using historical records to project hourly WBGT. Lastly, we analyze our WBGT projections to determine how future trends and extremes may affect outdoor training time at these locations.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData Sources\u003c/h2\u003e \u003cp\u003eAirfield weather stations at or near the study locations (Table S2) provide our hourly weather records for 01/01/1990 to 09/30/2022 including surface air temperature (\u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e in \u003csup\u003eo\u003c/sup\u003eC), surface pressure (in hPa), relative humidity (in g/kg), solar radiation (in W/m\u003csup\u003e2\u003c/sup\u003e), and 10m wind speed (in m/s). Missing values are interpolated as described in Patton and Doyle (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Continuous fixed point WBGT observations are rare (Takakura, Fujimori et al. \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) and, to our knowledge, no multi-decade data sets exist. Instead, we use hourly WBGT records for these locations obtained from the 14th WX, the organization providing authoritative weather and climate data to the U.S. military (HQDA \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), as historical observations.\u003c/p\u003e \u003cp\u003eGCM data sets are obtained from the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) (Thrasher 2021). We use the single grid cell at each location encompassing the weather station providing our historical records. Our methodology requires five input variables: temperature (\u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e), relative humidity, wind speed, solar radiation, and surface pressure. Of 35 GCMs included in the NEX-GDDP CMIP6 project, 23 include all required variables. After dropping one GCM known to have temperature deviations (Thrasher \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and three due to data access difficulties, we subject 19 to full analysis at each location (Table S3). Data sets are subdivided into three periods: 2000\u0026ndash;2014 (15 years) using GCM historical runs to evaluate the accuracy of our method against observations; 2015\u0026ndash;2021 (7 years) using GCM scenario runs compared against observations for bias adjustment; and 2025\u0026ndash;2100 (76 years) to project future WBGT at study locations.\u003c/p\u003e \u003cp\u003eWe select the two IPCC scenario Shared Socioeconomic Pathways (SSPs) that currently appear to reflect more plausible future outcomes (Hausfather and Peters \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2020\u003c/span\u003e): SSP 2-4.5 \u003cem\u003emiddle-of-the-road\u003c/em\u003e and SSP 3\u0026ndash;7.0 \u003cem\u003eregional rivalry\u003c/em\u003e (Hausfather \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). SSP 3 is becoming the choice scenario for planning in a high emission future in some impact assessments (NMFS \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and SSP 2 the most likely scenario given globally pledged climate policies (Scafetta \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In contrast, SSP 1 and SSP 5 scenarios appear increasingly unlikely (Kriegler, Bauer et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2017\u003c/span\u003e, Riahi, van Vuuren et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). SSP 2 and SSP 3 also bracket the likely range of warming (2.1-3.4\u003csup\u003eo\u003c/sup\u003eC, median 2.8\u003csup\u003eo\u003c/sup\u003eC) projected by 2100 (IPCC \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This likely warming range falls within the projected range for SSP 2 (2.1-3.5\u003csup\u003eo\u003c/sup\u003eC). SSP 3\u0026rsquo;s higher range (2.8-4.6\u003csup\u003eo\u003c/sup\u003eC) makes it suitable for lower-probability but higher-risk impact evaluation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eEvaluating the Liljegren method\u003c/h2\u003e \u003cp\u003eWe calculate WBGT as described by Liljegren (2008) using R package \u0026ldquo;wbgt\u0026rdquo; (Lieblich \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). This method is recommended in comparative studies (Lemke and Kjellstrom \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2012\u003c/span\u003e, Kong and Huber \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and implemented in similar work (Ahn, Uejio et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, Lewandowski, Kioumourtzoglou et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). We evaluate the method in three ways: (1) using hourly historical weather observations (2000\u0026ndash;2014) as inputs and comparing results to WBGT provided by the 14th WX, (2) correlating weather observations with daily maximum, mean, and minimum WBGT from the 14th WX and Liljegren methods, and (3) using steps #1\u0026ndash;3 of this study\u0026rsquo;s methodology (described below) to estimate WBGT over the same period using inputs from 19 GCM historical runs.\u003c/p\u003e \u003cp\u003eFinding the Liljegren method appropriate, we apply our novel methodology to estimate hourly WBGT at each location between 01/01/2025 and 12/31/2100. We use 5 steps: (1) finding hourly WBGT for a climatological average year, (2) estimating required weather observation values unavailable from GCM outputs, (3) creating uncorrected hourly WBGT by shifting the average WBGT diurnal curve to match WBGT values calculated from GCM outputs, (4) selecting a six model GCM subset based on a goodness-of-fit test evaluating model skill, and (5) bias adjusting the subset.\u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section3\"\u003e \u003ch2\u003eStep #1: WBGT values in an \u0026ldquo;average year\u0026rdquo;\u003c/h2\u003e \u003cp\u003eOur first step is finding mean values at each hour of a year for \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e, wind speed, relative humidity, solar radiation, and surface pressure using observations between 01/01/1990\u0026ndash;09/30/2022 and after removing leap days. The resulting data set contains 1 years\u0026rsquo; worth of mean hourly observations (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;8,760) for each variable. We consider these the variable climatological averages for their respective hour of the year.\u003c/p\u003e \u003cp\u003eThe Liljegren method can return unrealistic values if wind \u0026cong; 0 m/s. Previous WBGT studies used minimums of 0.5 m/s (Spangler, Liang et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) or 1.0 m/s (Lemke and Kjellstrom \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2012\u003c/span\u003e), or apply a correction factor to velocities below 0.5 m/s (Patel, Mullen et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), recognizing in \u0026ldquo;still air\u0026rdquo; body movement and natural processes generate some minimum airflow over skin. GCM wind speed output is provided at 10m height, so we estimate a speed of 0.62 m/s returning 0.5 m/s at 2m (Eq.\u0026nbsp;2) and adjust all observations\u0026thinsp;\u0026lt;\u0026thinsp;0.62 m/s to 0.62 m/s.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003eEquation 2: Windspeed height profile formula\u003c/h2\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\({v}_{2}={v}_{1}*\\left[\\right(\\text{ln}\\left(\\frac{{h}_{2}}{{z}_{0}}\\right))/(\\text{ln}\\left(\\frac{{h}_{1}}{{z}_{0}}\\right))\\)\u003c/span\u003e \u003c/span\u003e]\u003c/p\u003e \u003cp\u003eWhere: \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({v}_{x}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({h}_{x}\\)\u003c/span\u003e\u003c/span\u003eare the velocity (in m/s) of wind at height \u003cem\u003eh\u003c/em\u003e\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\({z}_{0}\\)\u003c/span\u003e \u003c/span\u003e is our chosen roughness length of 0.0024m\u003c/p\u003e \u003cp\u003eAfter adjusting wind speed, hourly WBGT is estimated for every hour between 01/01/1990\u0026ndash;09/30/2022. The mean WBGT for each hour across years is then calculated, establishing a climatological average WBGT value for each hour of the year. This climatological average WBGT, consisting of 365 averaged WBGT diurnal cycles, forms the baseline from which future hourly values are later derived.\u003c/p\u003e \u003cp\u003e \u003cb\u003eStep #2: Creating input variables: the Liljegren method and the NEX-GDDP data sets.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eHere we describe how the required input variables are obtained or estimated from GCM output. Our method requires daily maximum, mean, and minimum values for each input variable. The NEX-GDDP data sets provide six: daily max, mean, and min \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e, and daily mean wind speed, relative humidity, and solar radiation. Many NEX-GDDP GCM outputs do not include surface pressure, leaving nine missing values to be estimated: daily maximum and minimum for wind speed, solar radiation, and relative humidity, and daily maximum, mean, and minimum surface pressure. Relative humidity responds non-linearly with changing \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e, making it difficult to estimate maximum and minimum from a daily mean value. We choose instead to calculate relative humidity values from specific humidity (described below).\u003c/p\u003e \u003cp\u003eTo ensure projected WBGTs are mapped to the correct part of the diurnal cycle, the hour associated with minimum, mean, and maximum WBGT at each location and on each day of the year is identified from the 14th WX data set. This allows the time associated with maximum and minimum WBGT to change depending on seasonal and local considerations.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section3\"\u003e \u003ch2\u003e2.1 Estimating wind speed and solar radiation\u003c/h2\u003e \u003cp\u003eTo determine maximum and minimum wind speed we use a scaling factor unique to each location and day of the year. We find speed corresponding with the hour of daily maximum, mean, and minimum WBGT across fifteen years of observations (01/01/2000\u0026ndash;12/31/2014). The difference between max (min) and mean wind speed is calculated for each day. Differences are grouped by day to create a unique average daily scaling factor for mean-to-max and mean-to-min wind speed for each day and location. The same method is applied to estimate daily maximum solar radiation. Minimum solar radiation is fixed at 0 W/m\u003csup\u003e2\u003c/sup\u003e on the assumption that minimum WBGT occurs at night. To prevent outliers from skewing scaling factors in this relatively small sample (the mean of 15 observations creates each daily scaling factor), observations\u0026thinsp;\u0026gt;\u0026thinsp;99th and \u0026lt;\u0026thinsp;1st percentile are replaced with the 1st and 99th percentile.\u003c/p\u003e \u003cp\u003eThe result is three sets of 365 scaling factor values for each location: one each for the difference between maximum-to-mean and minimum-to-mean wind speed and one between maximum-to-mean solar radiation. Scaling factors are added to corresponding daily mean GCM values to estimate daily maximum and minimum values, inherently accounting for intra-year and seasonal changes. We correct the few unrealistic maximum solar radiation values by limiting solar radiation to each locations historical observed maximum. Daily maximum wind speed values greater than those in our historical data sets, also rare, are retained on the assumption that GCMs may predict future wind speeds above those previously recorded (Table S4).\u003c/p\u003e \u003cp\u003e \u003cem\u003e2.2 Estimating surface pressure.\u003c/em\u003e \u003c/p\u003e \u003cp\u003eAlthough similar studies have used the ideal gas law to estimate surface pressure based on temperature and elevation (Chavaillaz, Roy et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) our study found it performs poorly when compared against hourly observations. Instead, three linear models are developed at each study location using historical records (Eq.\u0026nbsp;3). These models correlate observed surface pressure at hours associated with WBGT maximum, mean, and minimum with the \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e, wind speed, relative humidity, and solar radiation observed at those same hours to estimate surface pressure distributions better matching observations.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eEquation 3: Surface Pressure Linear Model\u003c/h2\u003e \u003cp\u003e \u003cdiv id=\"Equb\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$Surface Pressure ={}_{1}\\left({T}_{a}\\right)+{}_{2}\\left(Wind Speed\\right)+{}_{3}\\left(Relative Humidity\\right)+{}_{4}\\left(Solar Radition\\right)+X$$\u003c/div\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003eTo remove any systemic bias in the distribution of maximum or minimum surface pressure, values are shifted by the difference between the mean observed and calculated maximum (minimum) surface pressure values, \u0026ldquo;nudging\u0026rdquo; the distribution to align more closely with observations (Fig. S5). These surface pressure values are location specific since models are developed from local weather records.\u003c/p\u003e\u003cp\u003e \u003cem\u003e2.3 Estimating relative humidity.\u003c/em\u003e \u003c/p\u003e \u003cp\u003eWe use daily mean specific humidity to calculate relative humidity. By assuming the mass of water vapor remains constant throughout the day, daily minimum, mean, and maximum relative humidity can be found as a function of temperature and pressure. We implement this calculation with R package \u0026lsquo;humidity\u0026rsquo; (Cai \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Inputs are daily mean specific humidity and daily max, mean, and minimum \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e (from GCMs), and daily max, mean, and min surface pressure.\u003c/p\u003e \u003cp\u003eSpecific humidity was not provided in the 14th WX data, preventing direct comparison between observed and calculated relative humidity using this method. As a proxy to test robustness of this method, relative humidity values from historical GCM runs are calculated and compared to observations between 2000\u0026ndash;2014. Calculated values are slightly overestimated at the hour of daily minimum WBGT, while distributions better match observations at mean and maximum WBGT. (Fig. S6).\u003c/p\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003eStep #3: Uncorrected Hourly Wet Bulb Globe Temperatures\u003c/h2\u003e \u003cp\u003eAfter obtaining all 15 daily input values, we calculate daily maximum, mean, and minimum WBGT per Liljegren (2008). To extend WBGT to hourly values, we shift the climatological average diurnal curves from step #1 to new positions by calculating the difference between the GCM-derived WBGT and the corresponding WBGT from the climatological average year. For example, GCM-derived WBGTs of 25\u003csup\u003eo\u003c/sup\u003eC (maximum), 20\u003csup\u003eo\u003c/sup\u003eC (mean), and 15\u003csup\u003eo\u003c/sup\u003eC (minimum) on May 1st, 2050, and climatological average WBGT on May 1st of 23\u003csup\u003eo\u003c/sup\u003eC, 20\u003csup\u003eo\u003c/sup\u003eC, and 12\u003csup\u003eo\u003c/sup\u003eC, result in differences associated with May 1st, 2050, of 2\u003csup\u003eo\u003c/sup\u003eC, 0\u003csup\u003eo\u003c/sup\u003eC, and 3\u003csup\u003eo\u003c/sup\u003eC (Table S7).\u003c/p\u003e \u003cp\u003eSince diurnal cycles form sinusoidal waves, mean differences occur twice daily. Each day\u0026rsquo;s differences are therefore associated with four hours- once at maximum WBGT, once at minimum, and twice at mean. For the remaining twenty hours the difference between GCM WBGT and climatological average WBGT is created by interpolating between these four differences. All differences are then added back to the climatological average WBGT, creating twenty new WBGT values and recreating the daily maximum, mean, and minimum differences. This is repeated for each day and model (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eStep #4: Selecting the best GCMs for each location\u003c/h2\u003e \u003cp\u003eWe select a subset of six GCMs per location, ensuring GCMs with the best local skill are used in subsequent analysis. We rank GCMs by comparing each model\u0026rsquo;s WBGT cumulative distribution function (CDF) from 15 years of historical runs (2000\u0026ndash;2014) to the WBGT CDF from the 14th WX data sets over the same period. Given our focus on high WBGT, comparing CDF tails (specifically the right \u0026ldquo;hot\u0026rdquo; tail) is more relevant than overall CDF shape in ranking GCMs. We use the two-sample Anderson-Darling (AD) test (Anderson and Darling \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1952\u003c/span\u003e) implemented with R package \u0026ldquo;twosamples\u0026rdquo; (Dowd \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This test considers difference in CDFs shape and symmetry and is more sensitive to differences at distribution tails than other tests (Engmann and Cousineau \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWe also compare WBGT provided by the 14th WX and WBGT calculated from weather observations using the Liljegren method (Fig. S8). Test statistics comparing the 14th WX data set with these WBGTs are much smaller than the test statistics between the 14th WX data set and any GCM CDF, providing additional evidence that the Liljegren method is robust for calculating WBGT since distributions closely align. Supplemental table S3 provides AD test statistics for each GCM output.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eStep #5: Empirical Quantile Mapping to correct GCM bias\u003c/h2\u003e \u003cp\u003eBias adjustment is applied to account for systematic bias in climate models, particularly in regional and extreme events studies (Jeon, Paciorek et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e, Qian and Chang \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Lehner, Nadeem et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). We use quantile delta mapping (QDM) implemented in R package \u0026ldquo;qmap\u0026rdquo; (Gudmundsson \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) because QDM outperforms simpler methods when adjusting distribution tails (Qian and Chang \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Lehner, Nadeem et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and has been used in studies of climate change related health outcomes (Jeon, Paciorek et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOur QDM implementation generates transformation functions unique to each GCM, scenario, and location between the observed and modelled WBGT CDFs (i.e., 14th WX data and GCM runs between 2015\u0026ndash;2021) at every 0.5th quantile. These transformation functions are then applied to projected WBGT derived from GCMs (2025\u0026ndash;2100). The 0.5 percentile step recognizes the small percentage of the CDF containing impactful (i.e., extreme) WBGT values and adjusts the CDF appropriately (Fig. S9). For example, at our coolest study location (FLW), significantly impactful WBGT\u0026thinsp;\u0026gt;\u0026thinsp;31.11\u003csup\u003eo\u003c/sup\u003eC occur\u0026thinsp;\u0026lt;\u0026thinsp;1% of the time. A coarser transformation step would not appropriately adjust such a small percentage of values.\u003c/p\u003e \u003cp\u003eAdjustments are made to each year\u0026rsquo;s CDF independently, retaining trends between sequential years. For example, the transformation is applied to WBGT values for 2025 to create a bias-adjusted 2025 data set, then reapplied to 2026 data and so on for subsequent years. This assumes CDFs for future year remain similarly shaped to the CDF derived from the 14th WX data but allows each year\u0026rsquo;s CDF to progressively shift warmer.\u003c/p\u003e \u003cp\u003eData sets are complete after bias adjustment. At each location our final data sets include scenario specific, bias corrected hourly WBGT from six GCMs for the periods of 2015\u0026ndash;2021 and 2025\u0026ndash;2100.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eHeat Wave and Trend Evaluation\u003c/h2\u003e \u003cp\u003eOur heat wave assessment criteria is a modification of the Russo et al (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) Heat Wave Magnitude Index (HWMI). Inclusion criteria is 6 or more hours per day \u0026ge;31.11\u003csup\u003eo\u003c/sup\u003eC WBGT for three or more consecutive days, reflecting our desire to evaluate heat waves for their impacts on outdoor labor. For nearly all outdoor labor and recreation, activity modification is recommended at or before 31.11\u003csup\u003eo\u003c/sup\u003eC (United States. Occupational and Health 2017, Roberts, Armstrong et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, HQDA \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and 6 hours of \u0026ldquo;lost\u0026rdquo; time (a quarter of the day) reflects the potential loss of a workday. We do not determine subheat wave magnitude but in the interest in assessing labor impacts, heat wave magnitude is the product of mean hours per day\u0026thinsp;\u0026ge;\u0026thinsp;31.11\u003csup\u003eo\u003c/sup\u003eC and the length of the heat wave in days; the resulting HWMI approximates the cumulative hours\u0026thinsp;\u0026ge;\u0026thinsp;31.11\u003csup\u003eo\u003c/sup\u003eC for each heat wave. Following Russo et al (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), we choose HWMI as the maximum heat wave in a year. For projections, this is the greatest HWMI from any of the six GCMs.\u003c/p\u003e \u003cp\u003eAll trends are calculated using a seasonal variation of Mann-Kendall analysis, a robust method testing for monotonic trends in time series data (Donald Meals \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). We implement this test using R package \u0026lsquo;EnvStats\u0026rsquo; (Steven P. Millard \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). To test trends in future weather variables we use an unadjusted multi-model mean of daily mean values from all 19 GCMs.\u003c/p\u003e \u003c/div\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eEvaluating calculated WBGT\u003c/h2\u003e \u003cp\u003eOur first evaluation compares WBGT calculated using the Liljegren method with values provided by the 14th WX. Since both are derived from the same observations the two data sets should closely match. We find distributions agree at all locations (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) and same-hour WBGT comparison have r\u003csup\u003e2\u003c/sup\u003e values between 0.987\u0026ndash;0.992 (Fig. S10). The next evaluation, correlating weather variable inputs and the 14th WX WBGT and Liljegren calculated WBGT values, find correlations nearly identical (Fig. S11). Comparing un-adjusted multi-model means using GCM historical runs between 2000\u0026ndash;2014 and the 14th WX data returns differences between \u0026minus;\u0026thinsp;1.29 to 0.87\u003csup\u003eo\u003c/sup\u003eC and multi-model median differences between \u0026minus;\u0026thinsp;2.20 to 0.88\u003csup\u003eo\u003c/sup\u003eC WBGT (Table S12). Without selecting GCMs subset based on local skill and before bias adjusting, estimating WBGT from GCM output appears accurate; mean and median differences are reduced to \u0026lt;\u0026thinsp;\u0026lt;\u0026thinsp;1\u003csup\u003eo\u003c/sup\u003eC when comparing the selected and bias adjusted GCM subset outputs (2015\u0026ndash;2021) for either scenario (Table S13).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eCorrelations between input variables and output WBGT\u003c/h2\u003e \u003cp\u003eConsidering the variables that influence WBGT provides insights to the local timing and conditions when high WBGT occurs. \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e and specific humidity always correlate strongly with WBGT at all locations (Table S14). Daily mean \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e, not maximum, is most strongly correlated to maximum WBGT at all locations while daily minimum \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e is always correlated most strongly with daily minimum WBGT. At southeast locations (FMGA and FJSC) humidity is the variable second most strongly correlated with maximum WBGT, while at FSOK humidity is third (behind minimum \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e)\u003c/em\u003e and at FLW it is fourth (behind maximum and minimum \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e).\u003c/p\u003e \u003cp\u003eWind speed and solar radiation are weakly to moderately correlated with WBGT. Wind speed is always negatively correlated and is weakest at FSOK, the location with highest average velocities. Daily maximum and mean solar radiation are always positively correlated with WBGT, with correlation stronger with daily maximum WBGT than daily minimum, expected since minimum WBGT occurs prior to sunrise.\u003c/p\u003e \u003cp\u003eVisualizing intra-year and inter-year trends show how weather variables may influence future WBGT (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e; quantified in Figs. S15a-S15d). Specific humidity and \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e are projected to increase, driving a commensurate increase to the WBGT, while wind speed trends are small. There are no inter-year trends to solar radiation, but the timing of peak daily mean solar radiation is interesting in that it only occurs near the summer solstice at FLW; at the Southeast locations it occurs around mid-May. Seasonally and related factors such as cloud cover, humidity, and aerosols affect inter-year amounts of solar radiation incident on the surface. Intra-year peak solar radiation timing is consistent with estimates from solar power planning calculators at nearby locations (SEL \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIntra-year weather variable trends differ among locations, and understanding the synergistic timing of these trends provides insight to when, and for how long, WBGT may be elevated at a location. At FSOK, specific humidity peaks between June and July, while mean \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e doesn\u0026rsquo;t peak until late July. The relative early increase and slow decline in humidity along with the later peak in \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e results in an extended period of locally elevated WBGT compared to other locations, with mean WBGT peaking around the first week of July and remaining generally flat compared to curves at other installations (bottom row, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). In contrast, mean WBGT peaks between 23\u0026ndash;29 July (FJSC) and 26\u0026ndash;31 July (FMGA) with specific humidity values peaking a few days prior and mean \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e about a week earlier. At these locations WBGT peaks shortly after temperature and humidity because wind speed decreases through mid-August. At our fourth location (FLW), specific humidity, \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e, and WBGT all peak near the same period (17\u0026ndash;22 July) resulting in a narrower window of elevated mean WBGT.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eFuture WBGT Distributions\u003c/h2\u003e \u003cp\u003eWBGT distribution displays surprising changes at some locations, particularly in the warmer SSP 3 scenario where a bimodal peak emerges at FJSC and FMGA in later years (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). This peak is less well defined at FSOK and does not appear at FLW. Given influence of specific humidity and \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e on WBGT, emerging peaks at the warm end of the distribution are likely related to interaction between these two variables. Dividing WBGT distribution by day and night demonstrates the bi-modal peaks originate during daylight hour (Fig. S16). The distribution of nighttime WBGT, less influenced by humidity trends since relative humidity \u0026cong; 100% is more common at night, generally retain their shape while shifting warmer.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eWBGT index trends\u003c/h2\u003e \u003cp\u003eDaily minimum WBGT is projected to increase faster than daily mean WBGT, which increase faster than daily maximum WBGT; nights are warming faster than days (Table\u0026nbsp;2). This is consistent with trend assessments of \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sub\u003e within the USA (Vanos, Kalkstein et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2015\u003c/span\u003e, USGCRP \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cem\u003eTable 2:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cem\u003eTrends as the multi-model mean for daily maximum, mean, and minimum WBGT per decade.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"640\" height=\"289\"\u003e\u003c/em\u003e\u003cbr\u003e\u003c/p\u003e\u003cp\u003eConsistent with scenario design (O\u0026rsquo;Neill, Kriegler et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), SSP 3 shows a continued rate of WBGT warming as the century progresses, while SSP 2 shows a positive but decreasing rate between the early and late periods. During the warm season (May-October) the rate of daily minimum WBGT increase is as high as 0.625\u003csup\u003eo\u003c/sup\u003eC decade\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in SSP 3 at FSOK and 0.465\u003csup\u003eo\u003c/sup\u003eC decade\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in SSP 2 at FLW. In contrast, the rate of increase for daily maximum WBGT is 0.451\u003csup\u003eo\u003c/sup\u003eC decade\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in SSP 3 at FSOK and 0.323\u003csup\u003eo\u003c/sup\u003eC decade\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in SSP 2 at FLW.\u003c/p\u003e \u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eIncrease in WBGT extremes and heatwave frequency\u003c/h2\u003e \u003cp\u003eWe use heat category 4 (\u0026ge; 31.11\u003csup\u003eo\u003c/sup\u003eC) to identify WBGT significant enough to cause severe labor impacts. This value exceeds recommended limits for acclimatized workers performing \u0026ldquo;light\u0026rdquo; work (NIOSH \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), the Occupational Safety Health Administration recommend significant rest periods when working above this threshold (United States. Occupational and Health 2017) and consideration should be given to rescheduling exertional outdoor military training (Office of the Surgeon OSG \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The American College of Sports Medicine recommends rescheduling athletic competitions well before this threshold and limiting training to less than 1 hour for low-risk individuals (Roberts, Armstrong et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOur results show significant increases in the frequently of hours exceeding 31.11\u003csup\u003eo\u003c/sup\u003eC (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). By 2039, depending on location, hours \u0026ge; 31.11\u003csup\u003eo\u003c/sup\u003eC increase between 22\u0026ndash;63% above a 2015\u0026ndash;2021 baseline. By mid-century the number of hours per year above this threshold approximately doubles (88\u0026ndash;151%), with variations emerging between scenarios. By 2099, hours \u0026ge; 31.11\u003csup\u003eo\u003c/sup\u003eC increase at least 143% in the SSP 2 scenario and up to 443% in SSP 3. By century\u0026rsquo;s end, at each location hours above this threshold occur at least 100% more frequently in SSP 3 than in SSP 2 compared to baseline (Table S17).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eOur HWMI is a proxy for hot hours on consecutive days and may be more relevant for outdoor labor since it approximates hours potentially lost to outdoor labor during a heatwave (Table S18). Increases in consecutive days primarily drives increasing HWMI. At FLW, the location with the lowest heat burden, heatwaves historically averaged a HWMI of 28 and a duration of ~\u0026thinsp;4 days long (i.e., ~\u0026thinsp;28 hours \u0026ge; 31.11\u003csup\u003eo\u003c/sup\u003eC over four consecutive days). Based on projections, by 2039 the HWMI increases to 39 with an average length of 5 days; by the 2050s, a HWMI of 45 with a length of 6 days. In the SSP 2 scenario, by 2099 heatwaves are on average more than 8 days long at FLW. At the other three locations, heatwave duration increases from about 4 days to, on average, 20 days by mid-century, driven by the increase in consecutive days \u0026ge; 31.11\u003csup\u003eo\u003c/sup\u003eC. Hours per day \u0026ge; 31.11\u003csup\u003eo\u003c/sup\u003eC also increase but never exceeds 11 hours per day in any projection, likely due to cooling at night.\u003c/p\u003e \u003cp\u003eWeeks long heatwaves will create significant challenges for outdoor labor, and particularly for outdoor labor conducted on fixed schedules. For instance, U.S. military recruit training follows a 10-week program, construction often consists of sequential activities, and agricultural work depends on specific growing cycles and seasonal timing. These types of activities will struggle under extended periods of activity-limiting heat. The most extreme heatwave predicted in any model and scenario occurs at FMGA near the end of the century. At 64 days (over 9 weeks) long, during which 707 hours (most daytime hours) are \u0026ge; 31.11\u003csup\u003eo\u003c/sup\u003eC, such an event represents less of a heatwave and more of a permanent shift toward higher summer WBGT and would require extreme adaptations to any schedule dependent outdoor labor program.\u003c/p\u003e \u003cp\u003eIn some cases, labor can shift to cooler nighttime periods, but this presents a new series of challenges, including additional health risks (He, Kim et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The lowest threshold in our study, WBGT \u0026ge; 25.56\u003csup\u003eo\u003c/sup\u003eC, still presents risk for EHI. Individuals predisposed to EHI and those from cooler climates are at increased risk during athletic training and competition even at this lower threshold (Roberts, Armstrong et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) while those performing heavy labor are recommended to take rest breaks that can exceed 50% of total labor time (United States. Occupational and Health 2017, HQDA \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The effect of temperature increases on workplace injury rates correlate more with overnight lows compared with daytime highs (McInnes, Akram et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eHistorically few nights at our locations had minimum temperature \u0026ge; 25.56\u003csup\u003eo\u003c/sup\u003eC and GCM historical runs average 1 or fewer such nights per year. Such infrequent occurrence won\u0026rsquo;t remain the norm. Daily minimum WBGT are projected to increase across all locations, especially FSOK. In the SSP 2 scenario, three locations project\u0026thinsp;~\u0026thinsp;5 nights per year with minimum WBGT \u0026ge; 25.56\u003csup\u003eo\u003c/sup\u003eC between 2050\u0026ndash;2100. In contrast, FSOK projections frequently indicate 20\u0026thinsp;+\u0026thinsp;days per year in the last two decades of the century.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eStudy Limitations\u003c/h2\u003e \u003cp\u003eWBGT index is sensitive to microclimates (Clark and Konrad \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Solar, wind, or humidity patterns may differ between nearby locations depending on local topography. Our observations come from a single weather station. Projected WBGT is thus qualified as accurate for these point locations, although trends in surrounding areas are likely similar. Using climatological average values (reference step #1) introduces uncertainty. The 95th CI for hourly mean values is generally small during warm months (\u0026lt;\u0026thinsp;2\u003csup\u003eo\u003c/sup\u003eC), but this uncertainty is carried forward to projected WBGT (Table S19). Our bias adjustment method assumes future WBGT distributions can be approximated from past distributions. Despite limitations, we believe this study provides a precise methodology for projecting WBGT at point locations from GCM data due to our explicit calculations of WBGT and use of long-term local averages.\u003c/p\u003e \u003cp\u003eTwo further challenges limit implement at large spatial scales. Large computational effort is needed; here, we propose no solutions, since the Liljegren method requires two iterative calculations for each WBGT value. The second is the requirement for historical observations. With minimal modification, this method can theoretically be used in the absence of historical weather station records by replacing records with reanalysis data to calculate historic WBGT. Local climatological average hourly WBGT could thus be estimated (reference step #1), allowing this method to be applied in the absence of historical WBGT data sets.\u003c/p\u003e \u003c/div\u003e"},{"header":"CONCLUSION","content":"\u003cp\u003eOur study provides a method to project WBGT from GCM outputs. We test a WBGT estimation method developed by Liljegren by comparing outputs with WBGT values provided by the 14th WX, then by using the method with historical GCM run outputs instead of weather observations. WBGT is projected using GCM outputs from the NASA NEX-GDDP project by calculating future WBGT at sub-daily resolution and adjusting historical climatological average diurnal cycles to match these values, thereby projecting future daily WBGT diurnal curves. We select GCMs and conduct bias adjustment against historical observations to ensure outputs are derived from GCMs with the best local skill and reflect the correct distribution of warm extreme values. The result at each location is hourly WBGT between 01/01/2025\u0026ndash;12/31/2100 for two SSP scenarios from 6 GCMs.\u003c/p\u003e \u003cp\u003eWe quantify trends to WBGT and weather variables that influence WBGT, finding WBGT overnight lows increasing fastest at all locations. Weather variables trend similarly across all locations with the timing of intra-year maximum and minimum unique to each location. We consider safety thresholds designed to mitigate EHI, finding the frequency of daytime hours above 31.11\u003csup\u003eo\u003c/sup\u003eC, the duration and intensity of heat waves, and the number of nights above 25.56\u003csup\u003eo\u003c/sup\u003eC projected to increase. By quantifying these trends at high spatiotemporal resolution, we hope to provide a methodology to inform locally relevant adaptations required to continue outdoor labor and training despite climate change.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eThe authors have no competing interests to declare that are relevant to the content of this article.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eFunding Acknowledgements\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003ePatton is funded by the Advanced Strategic Planning and Policy Program, Command and General Staff College, U.S. Army.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eData Availability and Transparency\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eData sets and code can be downloaded at: https://doi.org/10.7924/r4st7st5t\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eAuthor Contributions\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eAll authors contributed to the study. Methodology design was performed by Patton and Li. Data generation and initial analysis was performed by Patton. The first draft of the manuscript was written by Patton and all authors were involved in manuscript revisions. All authors read and approved the final manuscript. \u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAhn, Y., C. K. Uejio, J. Rennie and L. Schmit (2022). \u0026quot;Verifying Experimental Wet Bulb Globe Temperature Hindcasts Across the United States.\u0026quot; \u003cu\u003eGeoHealth\u003c/u\u003e \u003cstrong\u003e6\u003c/strong\u003e(4): e2021GH000527.\u003c/li\u003e\n\u003cli\u003eAlele, F. O., B. S. Malau-Aduli, A. E. O. Malau-Aduli and J. C. M (2020). \u0026quot;Epidemiology of Exertional Heat Illness in the Military: A Systematic Review of Observational Studies.\u0026quot; \u003cu\u003eInt J Environ Res Public Health\u003c/u\u003e \u003cstrong\u003e17\u003c/strong\u003e(19).\u003c/li\u003e\n\u003cli\u003eAnderson, T. W. and D. A. Darling (1952). \u0026quot;Asymptotic Theory of Certain \u0026quot;Goodness of Fit\u0026quot; Criteria Based on Stochastic Processes.\u0026quot; \u003cu\u003eThe Annals of Mathematical Statistics\u003c/u\u003e \u003cstrong\u003e23\u003c/strong\u003e(2): 193-212.\u003c/li\u003e\n\u003cli\u003eBlazejczyk, K., Y. Epstein, G. Jendritzky, H. Staiger and B. Tinz (2012). \u0026quot;Comparison of UTCI to selected thermal indices.\u0026quot; \u003cu\u003eInternational Journal of Biometeorology\u003c/u\u003e \u003cstrong\u003e56\u003c/strong\u003e(3): 515-535.\u003c/li\u003e\n\u003cli\u003eBrimicombe, C., C. H. B. Lo, F. Pappenberger, C. Di Napoli, P. Maciel, T. Quintino, R. Cornforth and H. L. Cloke (2023). \u0026quot;Wet Bulb Globe Temperature: Indicating Extreme Heat Risk on a Global Grid.\u0026quot; \u003cu\u003eGeoHealth\u003c/u\u003e \u003cstrong\u003e7\u003c/strong\u003e(2): e2022GH000701.\u003c/li\u003e\n\u003cli\u003eBudd, G. M. (2008). \u0026quot;Wet-bulb globe temperature (WBGT)--its history and its limitations.\u0026quot; \u003cu\u003eJ Sci Med Sport\u003c/u\u003e \u003cstrong\u003e11\u003c/strong\u003e(1): 20-32.\u003c/li\u003e\n\u003cli\u003eBuzan, J. R. and M. Huber (2020). \u0026quot;Moist Heat Stress on a Hotter Earth.\u0026quot; \u003cu\u003eAnnual Review of Earth and Planetary Sciences\u003c/u\u003e \u003cstrong\u003e48\u003c/strong\u003e(1): 623-655.\u003c/li\u003e\n\u003cli\u003eCai, J. (2019). humidity: Calculate Water Vapor Measures from Temperature and Dew Point. CRAN. \u003cstrong\u003e0.1.5\u003c/strong\u003e.\u003c/li\u003e\n\u003cli\u003eChavaillaz, Y., P. Roy, A.-I. Partanen, L. Da Silva, \u0026Eacute;. Bresson, N. Mengis, D. Chaumont and H. D. Matthews (2019). \u0026quot;Exposure to excessive heat and impacts on labour productivity linked to cumulative CO2 emissions.\u0026quot; \u003cu\u003eScientific Reports\u003c/u\u003e \u003cstrong\u003e9\u003c/strong\u003e(1): 13711.\u003c/li\u003e\n\u003cli\u003eClark, J. and C. E. Konrad (2024). \u0026quot;Observations and Estimates of Wet-Bulb Globe Temperature in Varied Microclimates.\u0026quot; \u003cu\u003eJournal of Applied Meteorology and Climatology\u003c/u\u003e \u003cstrong\u003e63\u003c/strong\u003e(2): 305-319.\u003c/li\u003e\n\u003cli\u003eDonald Meals, J. S., Steven Dressing, Jon Harcum (2011). Statistical Analysis for Monotonic Trends. \u003cu\u003eTech Notes\u003c/u\u003e, National Nonpoint Source Monitoring Program (NNPSMP).\u003c/li\u003e\n\u003cli\u003eDowd, C. (2023). twosample. CRAN.\u003c/li\u003e\n\u003cli\u003eEngmann, S. and D. Cousineau (2011). \u0026quot;Comparing distributions: the two-sample Anderson\u0026ndash;Darling test as an alternative to the Kolmogorov\u0026ndash;Smirnov test.\u0026quot; \u003cu\u003eJournal of Applied Quantitative Methods\u003c/u\u003e \u003cstrong\u003e6\u003c/strong\u003e: 1-17.\u003c/li\u003e\n\u003cli\u003eGudmundsson, L. (2016). Statistical Transformations for Post-Processing Climate Model Output\u003cstrong\u003e: \u003c/strong\u003eEmpirical adjustment of the distribution of variables originating from (regional) climate model simulations using quantile mapping.\u003c/li\u003e\n\u003cli\u003eHausfather, Z. (2018). \u0026quot;Explainer: How \u0026lsquo;Shared Socioeconomic Pathways\u0026rsquo; explore future climate change.\u0026quot; Retrieved February 2nd,, 2024, from https://www.carbonbrief.org/explainer-how-shared-socioeconomic-pathways-explore-future-climate-change/.\u003c/li\u003e\n\u003cli\u003eHausfather, Z. and G. P. Peters (2020). \u0026quot;Emissions - the \u0026apos;business as usual\u0026apos; story is misleading.\u0026quot; \u003cu\u003eNature\u003c/u\u003e \u003cstrong\u003e577\u003c/strong\u003e(7792): 618-620.\u003c/li\u003e\n\u003cli\u003eHe, C., H. Kim, M. Hashizume, W. Lee, Y. Honda, S. E. Kim, P. L. Kinney, A. Schneider, Y. Zhang, Y. Zhu, L. Zhou, R. Chen and H. Kan (2022). \u0026quot;The effects of night-time warming on mortality burden under future climate change scenarios: a modelling study.\u0026quot; \u003cu\u003eThe Lancet Planetary Health\u003c/u\u003e \u003cstrong\u003e6\u003c/strong\u003e(8): e648-e657.\u003c/li\u003e\n\u003cli\u003eHosokawa, Y., D. J. Casa, J. M. Trtanj, L. N. Belval, P. A. Deuster, S. M. Giltz, A. J. Grundstein, M. D. Hawkins, R. A. Huggins, B. Jacklitsch, J. F. Jardine, H. Jones, J. B. Kazman, M. E. Reynolds, R. L. Stearns, J. K. Vanos, A. L. Williams and W. J. Williams (2019). \u0026quot;Activity modification in heat: critical assessment of guidelines across athletic, occupational, and military settings in the USA.\u0026quot; \u003cu\u003eInternational journal of biometeorology\u003c/u\u003e \u003cstrong\u003e63\u003c/strong\u003e(3): 405-427.\u003c/li\u003e\n\u003cli\u003eHQDA (2021). AR 115-10/AFI 15-157 Weather Support for the U.S. Army. Washington DC, Department of the Army and Department of the Air Force.\u003c/li\u003e\n\u003cli\u003eHQDA (2022). TB MED 507: Heat Stress Control and Heat Casualty Management. Washington DC, Headquarters, Department of the Army.\u003c/li\u003e\n\u003cli\u003eIPCC (2021). Sixth Assessment Report Working Group II - Impacts, Adaptation and Vulnerability. \u003cu\u003eSixth Assessment Report\u003c/u\u003e, Intergovernmental Panel on Climate Change.\u003c/li\u003e\n\u003cli\u003eIPCC (2023). Climate Change 2023: Synthesis Report. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. J. R. H. Lee. Geneva, Switzerland.\u003c/li\u003e\n\u003cli\u003eJeon, S., C. J. Paciorek and M. F. Wehner (2016). \u0026quot;Quantile-based bias correction and uncertainty quantification of extreme event attribution statements.\u0026quot; \u003cu\u003eWeather and Climate Extremes\u003c/u\u003e \u003cstrong\u003e12\u003c/strong\u003e: 24-32.\u003c/li\u003e\n\u003cli\u003eKnutson, T. R. and J. J. Ploshay (2016). \u0026quot;Detection of anthropogenic influence on a summertime heat stress index.\u0026quot; \u003cu\u003eClimatic Change\u003c/u\u003e \u003cstrong\u003e138\u003c/strong\u003e(1): 25-39.\u003c/li\u003e\n\u003cli\u003eKong, Q. and M. Huber (2022). \u0026quot;Explicit Calculations of Wet-Bulb Globe Temperature Compared With Approximations and Why It Matters for Labor Productivity.\u0026quot; \u003cu\u003eEarth\u0026apos;s Future\u003c/u\u003e \u003cstrong\u003e10\u003c/strong\u003e(3): e2021EF002334.\u003c/li\u003e\n\u003cli\u003eKriegler, E., N. Bauer, A. Popp, F. Humpen\u0026ouml;der, M. Leimbach, J. Strefler, L. Baumstark, B. L. Bodirsky, J. Hilaire, D. Klein, I. Mouratiadou, I. Weindl, C. Bertram, J.-P. Dietrich, G. Luderer, M. Pehl, R. Pietzcker, F. Piontek, H. Lotze-Campen, A. Biewald, M. Bonsch, A. Giannousakis, U. Kreidenweis, C. M\u0026uuml;ller, S. Rolinski, A. Schultes, J. Schwanitz, M. Stevanovic, K. Calvin, J. Emmerling, S. Fujimori and O. Edenhofer (2017). \u0026quot;Fossil-fueled development (SSP5): An energy and resource intensive scenario for the 21st century.\u0026quot; \u003cu\u003eGlobal Environmental Change\u003c/u\u003e \u003cstrong\u003e42\u003c/strong\u003e: 297-315.\u003c/li\u003e\n\u003cli\u003eLehner, F., I. Nadeem and H. Formayer (2023). \u0026quot;Evaluating skills and issues of quantile-based bias adjustment for climate change scenarios.\u0026quot; \u003cu\u003eAdv. Stat. Clim. Meteorol. Oceanogr.\u003c/u\u003e \u003cstrong\u003e9\u003c/strong\u003e(1): 29-44.\u003c/li\u003e\n\u003cli\u003eLemke, B. and T. Kjellstrom (2012). \u0026quot;Calculating Workplace WBGT from Meteorological Data: A Tool for Climate Change Assessment.\u0026quot; \u003cu\u003eIndustrial Health\u003c/u\u003e \u003cstrong\u003e50\u003c/strong\u003e(4): 267-278.\u003c/li\u003e\n\u003cli\u003eLewandowski, S. A., M.-A. Kioumourtzoglou and J. L. Shaman (2022). \u0026quot;Heat stress illness outcomes and annual indices of outdoor heat at U.S. Army installations.\u0026quot; \u003cu\u003ePLOS ONE\u003c/u\u003e \u003cstrong\u003e17\u003c/strong\u003e(11): e0263803.\u003c/li\u003e\n\u003cli\u003eLi, C., X. Zhang, F. Zwiers, Y. Fang and A. M. Michalak (2017). \u0026quot;Recent Very Hot Summers in Northern Hemispheric Land Areas Measured by Wet Bulb Globe Temperature Will Be the Norm Within 20 Years.\u0026quot; \u003cu\u003eEarth\u0026apos;s Future\u003c/u\u003e \u003cstrong\u003e5\u003c/strong\u003e(12): 1203-1216.\u003c/li\u003e\n\u003cli\u003eLieblich, M. (2017). mdljts/wbgt. GitHub, University of Washington.\u003c/li\u003e\n\u003cli\u003eLiljegren, J. C., R. A. Carhart, P. Lawday, S. Tschopp and R. Sharp (2008). \u0026quot;Modeling the wet bulb globe temperature using standard meteorological measurements.\u0026quot; \u003cu\u003eJ Occup Environ Hyg\u003c/u\u003e \u003cstrong\u003e5\u003c/strong\u003e(10): 645-655.\u003c/li\u003e\n\u003cli\u003eMcInnes, J. A., M. Akram, E. M. MacFarlane, T. Keegel, M. R. Sim and P. Smith (2017). \u0026quot;Association between high ambient temperature and acute work-related injury: a case-crossover analysis using workers\u0026apos; compensation claims data.\u0026quot; \u003cu\u003eScand J Work Environ Health\u003c/u\u003e \u003cstrong\u003e43\u003c/strong\u003e(1): 86-94.\u003c/li\u003e\n\u003cli\u003eMinard, D. (1961). \u0026quot;Prevention of Heat Casualties in Marine Corps Recruits: Period of 1955\u0026ndash;60, With Comparative Incidence Rates And Climatic Heat Stresses in Other Training Categories.\u0026quot; \u003cu\u003eMilitary Medicine\u003c/u\u003e \u003cstrong\u003e126\u003c/strong\u003e(4): 261-272.\u003c/li\u003e\n\u003cli\u003eNIOSH (2016). Occupational Exposure to Heat and Hot Environments Revised Criteria 2016. D. o. H. a. H. Services. Concinnati, OH, Centers for Disease Control and Prevention, National Insititude for Occupational Safety and Health.\u003c/li\u003e\n\u003cli\u003eNMFS (2023). NATIONAL MARINE FISHERIES SERVICE PROCEDURE 02-110-18. DoC. USA, National Marine Fisheries Service, Department of Commerce.\u003c/li\u003e\n\u003cli\u003eO\u0026rsquo;Neill, B. C., E. Kriegler, K. L. Ebi, E. Kemp-Benedict, K. Riahi, D. S. Rothman, B. J. van Ruijven, D. P. van Vuuren, J. Birkmann, K. Kok, M. Levy and W. Solecki (2017). \u0026quot;The roads ahead: Narratives for shared socioeconomic pathways describing world futures in the 21st century.\u0026quot; \u003cu\u003eGlobal Environmental Change\u003c/u\u003e \u003cstrong\u003e42\u003c/strong\u003e: 169-180.\u003c/li\u003e\n\u003cli\u003eOSG (2023). Heat Illness Prevention for the 2023-2024 heat season. Washington, DC, Office of the Surgeon General, Department of the Army.\u003c/li\u003e\n\u003cli\u003eParsons, K. (2006). \u0026quot;Heat stress standard ISO 7243 and its global application.\u0026quot; \u003cu\u003eInd Health\u003c/u\u003e \u003cstrong\u003e44\u003c/strong\u003e(3): 368-379.\u003c/li\u003e\n\u003cli\u003eParsons, L. A., D. Shindell, M. Tigchelaar, Y. Zhang and J. T. Spector (2021). \u0026quot;Increased labor losses and decreased adaptation potential in a warmer world.\u0026quot; \u003cu\u003eNat Commun\u003c/u\u003e \u003cstrong\u003e12\u003c/strong\u003e(1): 7286.\u003c/li\u003e\n\u003cli\u003ePatel, T., S. P. Mullen and W. R. Santee (2013). \u0026quot;Comparison of Methods for Estimating Wet-Bulb Globe Temperature Index From Standard Meteorological Measurements.\u0026quot; \u003cu\u003eMilitary Medicine\u003c/u\u003e \u003cstrong\u003e178\u003c/strong\u003e(8): 926-933.\u003c/li\u003e\n\u003cli\u003ePatton, E. M. and M. W. Doyle (2023). \u0026quot;Observed Warming Trends at U.S. Army Basic Combat Training Installations and Implications for Future Recruit Training.\u0026quot; \u003cu\u003eMil Med\u003c/u\u003e.\u003c/li\u003e\n\u003cli\u003eP\u0026eacute;riard, J. D., D. DeGroot and O. Jay (2022). \u0026quot;Exertional heat stroke in sport and the military: epidemiology and mitigation.\u0026quot; \u003cu\u003eExp Physiol\u003c/u\u003e \u003cstrong\u003e107\u003c/strong\u003e(10): 1111-1121.\u003c/li\u003e\n\u003cli\u003ePisaniello, D. and R. Di Corleto (2023). Chapter 6 - Heat exposure and occupational health. \u003cu\u003eHeat Exposure and Human Health in the Context of Climate Change\u003c/u\u003e. Y. Guo and S. Li, Elsevier\u003cstrong\u003e: \u003c/strong\u003e109-131.\u003c/li\u003e\n\u003cli\u003eQian, W. and H. H. Chang (2021). \u0026quot;Projecting Health Impacts of Future Temperature: A Comparison of Quantile-Mapping Bias-Correction Methods.\u0026quot; \u003cu\u003eInt J Environ Res Public Health\u003c/u\u003e \u003cstrong\u003e18\u003c/strong\u003e(4).\u003c/li\u003e\n\u003cli\u003eRiahi, K., D. P. van Vuuren, E. Kriegler, J. Edmonds, B. C. O\u0026rsquo;Neill, S. Fujimori, N. Bauer, K. Calvin, R. Dellink, O. Fricko, W. Lutz, A. Popp, J. C. Cuaresma, S. Kc, M. Leimbach, L. Jiang, T. Kram, S. Rao, J. Emmerling, K. Ebi, T. Hasegawa, P. Havlik, F. Humpen\u0026ouml;der, L. A. Da Silva, S. Smith, E. Stehfest, V. Bosetti, J. Eom, D. Gernaat, T. Masui, J. Rogelj, J. Strefler, L. Drouet, V. Krey, G. Luderer, M. Harmsen, K. Takahashi, L. Baumstark, J. C. Doelman, M. Kainuma, Z. Klimont, G. Marangoni, H. Lotze-Campen, M. Obersteiner, A. Tabeau and M. Tavoni (2017). \u0026quot;The Shared Socioeconomic Pathways and their energy, land use, and greenhouse gas emissions implications: An overview.\u0026quot; \u003cu\u003eGlobal Environmental Change\u003c/u\u003e \u003cstrong\u003e42\u003c/strong\u003e: 153-168.\u003c/li\u003e\n\u003cli\u003eRoberts, W. O., L. E. Armstrong, M. N. Sawka, S. W. Yeargin, Y. Heled and F. G. O\u0026apos;Connor (2021). \u0026quot;ACSM Expert Consensus Statement on Exertional Heat Illness: Recognition, Management, and Return to Activity.\u0026quot; \u003cu\u003eCurr Sports Med Rep\u003c/u\u003e \u003cstrong\u003e20\u003c/strong\u003e(9): 470-484.\u003c/li\u003e\n\u003cli\u003eRusso, S., A. Dosio, R. G. Graversen, J. Sillmann, H. Carrao, M. B. Dunbar, A. Singleton, P. Montagna, P. Barbola and J. V. Vogt (2014). \u0026quot;Magnitude of extreme heat waves in present climate and their projection in a warming world.\u0026quot; \u003cu\u003eJournal of Geophysical Research Atmospheres\u003c/u\u003e \u003cstrong\u003e119\u003c/strong\u003e(22): 12,500-512,512.\u003c/li\u003e\n\u003cli\u003eSawka, M. N., L. R. Leon, S. J. Montain and L. A. Sonna (2011). \u0026quot;Integrated physiological mechanisms of exercise performance, adaptation, and maladaptation to heat stress.\u0026quot; \u003cu\u003eCompr Physiol\u003c/u\u003e \u003cstrong\u003e1\u003c/strong\u003e(4): 1883-1928.\u003c/li\u003e\n\u003cli\u003eScafetta, N. (2024). \u0026quot;Impacts and risks of \u0026ldquo;realistic\u0026rdquo; global warming projections for the 21st century.\u0026quot; \u003cu\u003eGeoscience Frontiers\u003c/u\u003e \u003cstrong\u003e15\u003c/strong\u003e(2): 101774.\u003c/li\u003e\n\u003cli\u003eSEL. (2024). \u0026quot;Solar Energy \u0026amp; Solar Power Resources.\u0026quot; Retrieved February 16th, 2024, from https://www.solarenergylocal.com.\u003c/li\u003e\n\u003cli\u003eSpangler, K. R., S. Liang and G. A. Wellenius (2022). \u0026quot;Wet-Bulb Globe Temperature, Universal Thermal Climate Index, and Other Heat Metrics for US Counties, 2000\u0026ndash;2020.\u0026quot; \u003cu\u003eScientific Data\u003c/u\u003e \u003cstrong\u003e9\u003c/strong\u003e(1): 326.\u003c/li\u003e\n\u003cli\u003eSteven P. Millard, A. K. (2022). EnvStats: Package for Environmental Statistics, Including US EPA Guidance.\u003c/li\u003e\n\u003cli\u003eTakakura, J., S. Fujimori, K. Takahashi, Y. Hijioka and Y. Honda (2019). \u0026quot;Site-specific hourly resolution wet bulb globe temperature reconstruction from gridded daily resolution climate variables for planning climate change adaptation measures.\u0026quot; \u003cu\u003eInt J Biometeorol\u003c/u\u003e \u003cstrong\u003e63\u003c/strong\u003e(6): 787-800.\u003c/li\u003e\n\u003cli\u003eThrasher, B. (2023). NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP-CMIP6) Technical Note, NASA.\u003c/li\u003e\n\u003cli\u003eThrasher, B., Wang, W., Michaelis, A. Nemani, R. (2021). NEX-GDDP-CMIP6. NASA. Center for Climate Simulation.\u003c/li\u003e\n\u003cli\u003eUnited States. Occupational, S. and A. Health (2017). \u003cu\u003eOSHA technical manual\u003c/u\u003e. Washington, D.C., U.S. Dept. of Labor, Occupational Safety and Health Administration.\u003c/li\u003e\n\u003cli\u003eUSGCRP (2023). \u003cu\u003eFifth National Climate Assessment\u003c/u\u003e. Washington, DC, USA.\u003c/li\u003e\n\u003cli\u003eVanos, J. K., L. S. Kalkstein and T. J. Sanford (2015). \u0026quot;Detecting synoptic warming trends across the US Midwest and implications to human health and heat-related mortality.\u0026quot; \u003cu\u003eInternational Journal of Climatology\u003c/u\u003e \u003cstrong\u003e35\u003c/strong\u003e(1): 85-96.\u003c/li\u003e\n\u003cli\u003eWillett, K. M. and S. Sherwood (2012). \u0026quot;Exceedance of heat index thresholds for 15 regions under a warming climate using the wet-bulb globe temperature.\u0026quot; \u003cu\u003eInternational Journal of Climatology\u003c/u\u003e \u003cstrong\u003e32\u003c/strong\u003e(2): 161-177.\u003c/li\u003e\n\u003cli\u003eYeargin, S., R. Hirschhorn, A. Grundstein, D. Arango, A. Graham, A. Krebs and S. Turner (2023). \u0026quot;Variations of wet-bulb globe temperature across high school athletics in South Carolina.\u0026quot; \u003cu\u003eInternational Journal of Biometeorology\u003c/u\u003e.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-biometeorology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"ijbm","sideBox":"Learn more about [International Journal of Biometeorology](http://link.springer.com/journal/484)","snPcode":"484","submissionUrl":"https://www.editorialmanager.com/ijbm/default2.aspx","title":"International Journal of Biometeorology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Climate change, wet bulb globe temperature, heat illness, outdoor labor","lastPublishedDoi":"10.21203/rs.3.rs-4414813/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4414813/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Increasing temperature will impact future outdoor worker safety but quantifying this impact to develop local adaptations is challenging. Wet bulb globe temperature (WBGT) is the preferred thermal index for regulating outdoor activities in occupational health, athletic, and military settings, but global circulation models (GCMs) have coarse spatiotemporal resolution and do not always provide outputs required to project the full diurnal range of WBGT. This article presents a novel method to project WBGT at local spatial and hourly temporal resolutions without many assumptions inherent in previous research. We calculate sub-daily future WBGT from GCM output and then estimate hourly WBGT based on a site-specific, historical diurnal cycles. We test this method against observations at U.S. Army installations and find results match closely. We then project hourly WBGT at these locations from January 1, 2025, to December 31, 2100, to quantify trends and estimate future periods exceeding outdoor activity modification thresholds. We find regional patterns affecting WBGT, suggesting accurately projecting WBGT demands a localized approach. Results show increased frequency of hours at high WBGT and, using U.S. military heat thresholds, we estimate impacts to future outdoor labor. By mid-century, some locations are projected to experience an average of 20 or more days each summer when outdoor labor will be significantly impacted. The method’s fine spatiotemporal resolution enables detailed analysis of WBGT projections, making it useful applied at specific locations of interest.","manuscriptTitle":"Wet bulb globe temperature from climate model outputs: a method for projecting hourly site-specific values and trends.","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-06-04 11:53:42","doi":"10.21203/rs.3.rs-4414813/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2024-05-22T19:45:02+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-05-22T16:40:20+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-05-16T12:42:36+00:00","index":"","fulltext":""},{"type":"submitted","content":"International Journal of Biometeorology","date":"2024-05-16T08:26:35+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-biometeorology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"ijbm","sideBox":"Learn more about [International Journal of Biometeorology](http://link.springer.com/journal/484)","snPcode":"484","submissionUrl":"https://www.editorialmanager.com/ijbm/default2.aspx","title":"International Journal of Biometeorology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"dd10ed08-fa4a-476c-b2c1-63f372ddfcd5","owner":[],"postedDate":"June 4th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-09-23T16:01:14+00:00","versionOfRecord":{"articleIdentity":"rs-4414813","link":"https://doi.org/10.1007/s00484-024-02776-5","journal":{"identity":"international-journal-of-biometeorology","isVorOnly":false,"title":"International Journal of Biometeorology"},"publishedOn":"2024-09-17 15:57:20","publishedOnDateReadable":"September 17th, 2024"},"versionCreatedAt":"2024-06-04 11:53:42","video":"","vorDoi":"10.1007/s00484-024-02776-5","vorDoiUrl":"https://doi.org/10.1007/s00484-024-02776-5","workflowStages":[]},"version":"v1","identity":"rs-4414813","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4414813","identity":"rs-4414813","version":["v1"]},"buildId":"7rjqhiLT3MXkJMwkYKINL","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}