Evaluating Heat Risk: Comparing On-Site WBGT Measurements Versus Smartphone Application Estimates

Abstract

Exertional heat illness poses a significant risk for workers, athletes, and military personnel participating in outdoor activities during hot weather. An important component of heat safety is to monitor environmental conditions through heat stress indices like the wet bulb globe temperature (WBGT) and adjust activity as conditions get progressively hotter. Traditionally, on-site (OS) WBGT measurement devices are used, but phone applications (PAs) offering WBGT estimates have emerged as a potential alternative. However, there is little information on how closely PA-derived WBGTs match OS measurements to guide decision-making. This study compared the PA-derived Zelus WBGT estimates with OS measurements from Kestrel 5400 devices and their impact on activity modification categorization. A 2-month observational study collected 1,056 paired (OS and PA) WBGT measurements from 26 high schools across 11 states in the United States and over diverse surfaces (artificial turf 53%, natural grass 44%, others 3%). WBGT values were categorized using regional activity modification thresholds to account for local acclimatization. Our findings indicated that while exhibiting high correlation (r = 0.89), PA WBGTs were on average about 1°C cooler, with differences of 2-3°C at higher WBGTs. Findings were similar for both grass and artificial turf surfaces. Further, significant discrepancies were observed in WBGT-based activity modification categories, with the PA more frequently indicating lower modification categories compared to OS devices, especially in hotter conditions. In light of these findings, the PA requires further validation prior to its adoption as a replacement for OS measurements.

Keywords: WBGT; activity modification; exertional heat illness; on‐site measurement; phone application; wet bulb globe temperature.

Figure 1

Scatter plots of phone application (PA). v. on‐site (OS) data. (a) wet bulb globe temperature (WBGT), (b) wet bulb temperature (WB), (c) globe temperature (GT), (d) dry bulb temperature (DB). The solid black line is the one‐to‐one line.

Figure 2

Forest plot of mean phone application (PA) and on‐site (OS) differences (PA‐OS) with 95% confidence intervals of wet bulb globe temperature (WBGT), natural wet bulb (NWB), globe temperature (GT), and dry bulb temperature (DB) over all surfaces and by surface type. All is all data, AT is artificial turf, and NG is natural grass.

Figure 3

On‐site (OS) versus phone application (PA) difference by OS WBGT. A negative difference indicates that the OS is greater than the PA WBGT. In the box plot, the lower and upper bounds of the box represent the 25th and 75th percentiles respectively with the black horizontal line the median, the lower and upper whiskers are the 10th and 90th percentiles respectively, and the lower and upper outliers represent the 5th and 95th percentile values respectively.

Figure 4

Frequency chart of regional Georgia High School Association (GHSA) activity modification categories for phone application (PA) and on‐site (OS) WBGT values. The categories range from 1 (no modification) to category 5 (cancellation of activities). Table S2 in Supporting Information S1 provides details on the activity modification categories. There is a significant difference between regional WBGT‐based activity modification indicated by OS compared to PA, χ² (4, N = 1,056) = 97.2174, p < 0.00001.