A dual-index assessment of warm-season heat stress for a small-island province in Atlantic Canada

Abstract Heat-stress exposure in maritime regions reflects the combined influence of temperature and humidity, yet these relationships remain poorly characterized in small, cool-temperate regions such as Prince Edward Island (PEI), Canada. Here, we characterize warm-season thermal exposure across PEI during a 2020–2025 observational baseline using Humidex, representing human thermal discomfort, and the Temperature–Humidity Index (THI), representing cattle heat-stress risk. Quality-controlled station observations were aggregated into a PEI-wide daily series, with daily mean temperature and relative humidity used as the primary exposure definition and daily maximum temperature evaluated separately as a sensitivity analysis. We show that relative humidity remained high throughout the warm season but acted primarily as a modifier of heat-stress thresholds, while temperature accounted for most day-to- day variability in both indices. Under daily mean conditions, Humidex ≥30 and THI ≥72 each occurred on 54 of 918 warm-season days, with the two classifications agreeing on 99.8% of days; no days reached Humidex ≥40 or THI ≥79. Most THI ≥72 exposure was nevertheless persistent, with 41 of 54 exceedance days occurring within 13 episodes lasting at least two consecutive days. Replacing daily mean with maximum temperature increased the number of Humidex ≥30 and THI ≥72 days to 350 and 346, respectively. These findings show that persistent humidity can narrow the temperature margin for elevated thermal exposure without producing continuously high daily mean heat stress, while also demonstrating that exposure estimates derived from daily meteorological data are highly sensitive to the representation of temperature.