Spatial-temporal insights into heat stress metrics for regional heat hazard and comfort assessment

The increasing frequency of extreme heat events due to climate change calls for a deeper understanding of the factors contributing to heat stress-related morbidity and mortality. Urban areas, with their high population densities, are especially vulnerable due to the urban heat island (UHI) effect, which intensifies temperatures compared to rural surroundings. Traditional heat stress assessments often focus solely on air temperature (Ta), overlooking other factors such as humidity, wind speed, and solar radiation important for the thermal comfort. Spatial data on heat stress is often lacking at fine spatial resolutions, making the use of LST a common. However the relationship between Ta, LST and thermal comfort metrics such as Universal Thermal Climate Index (UTCI) or Humidex are not yet well understood.  
To address this gap, we conducted a spatio-temporal assessment of thermal comfort metrics, LST and Ta across Hesse, Germany, using 1 km scale data. By examining temperature anomalies (ΔT), the difference between local and comparable rural background temperatures, we quantify the urban impact on heat stress while minimizing broader climatic influences. Diurnal and spatial patterns of temperature variations and thermal conditions are examined across different land use types and urban forms. Satellite-derived parameters were also incorporated to assess regional heat risk in areas lacking local measurement data.  
Our findings reveal a stronger UHI effect in ΔHumidex (max. 4.3°C) compared to ΔTa (max. 2.9°C) and ΔLST (max. 3.4°C), indicating that reliance on LST or Ta alone might underestimates the full extent of heat stress of the urban population. This is particularly significant as nearly one-third of the population in Hesse (30.4% for Ta, 25.6% for Humidex, and 34.7% for LST) lives in areas where temperatures exceed baseline levels both during the day and at night. The duration of temperature exceedance is consistently longer in urban areas, with average values of 3.8 hours and 2.8 hours of the diurnal cycle for ΔHumidex and ΔTa, respectively, compared to 0.7 hours and 0.4 hours in rural areas. Densely built areas, where nighttime cooling deficits persist longer, are particularly vulnerable, while inhabitants of open urban areas experience more moderate heat stress. These results emphasize the need to incorporate thermal comfort metrics that account for multiple parameters when evaluating the UHI effect. Although LST is commonly used as a substitute in UHI studies, its correlation with Ta and Humidex varies both spatially and temporally, warranting cautious application.