Assessment of thermal indices for the operational evaluation of thermal conditions in Athens, Greece

Thermal indices are valuable tools for assessing thermal environments linked to applications related to architecture, planning, tourism, energy conservation and health. They combine meteorological variables into a single value which can be assigned to a category of an assessment scale expressing the predicted degree of human thermal discomfort, sensation or stress. The assessment scales include one indifference (neutral) category and negative and/or positive categories of increasing intensity of thermal feeling. The numerous indices that have been developed makes the selection of a suitable one for a specific application a complex issue. The aim of this study was to assess the effectiveness of several commonly used thermal indices for their operational use in weather applications in the Mediterranean climate of Greece. Hourly data (2010-2021) of air temperature (Tair, ^oC), relative humidity (Rh, %), wind speed (WS, m/s) and global solar radiation (SR, W/m²) recorded in 15 weather stations across the Athens metropolitan area of the Automatic Weather Stations Network of the National Observatory of Athens were used to calculate the indices. The indices should follow the following criteria: (a) be currently in common use, (b) be easily estimated operationally, i.e., short estimation time using Tair, Rh, WS and SR, and (c) provide exploitable results. Apparent temperature (AT), Heat Index (HI), Humidex (HU), Normal Effective Temperature (NET), Physiologically Equivalent Temperature (PET), Universal Thermal Climate Index (UTCI), Wet-Bulb Globe Temperature (WBGT) and Wind Chill Index (WCI) were identified as the most commonly used thermal indices in research and by weather services around the world. PET showed the highest computational demand compared to the other indices. At the first step, NET, PET and UTCI were assessed as more suitable for the climate of Athens (compared to AT, HI, HU, WBGT and WCI), extending their predictions to the entire range of their assessment scales. NET and PET tended to classify more often than UTCI thermal conditions in the negative categories of their assessment scales (67.8%-NET and 62.1%-PET versus 36.8%-UTCI; p<0.001) while most of UTCI values were classified in the neutral category (48.9%-UTCI versus 15.6%-NET and 15.1%-PET; p<0.001). Finally, more heat wave days were classified in accordance to Tair by NET (86.8%) and UTCI (84.5%) compared to PET (70.5%; p=0.014). According to the results of this study, NET and UTCI could be suggested as the best candidate indices for operational use in the climate of Athens, Greece.