A systematic review advocating a framework and benchmarks for assessing outdoor human thermal perception.

Since the early 2000's, much attention has been paid to human thermal assessment in urban outdoor environments in different climatic zones. Some previous studies have argued that an absence of an agreed protocol for outdoor human bio-meteorological research causes complexity in comparing the studies' results for several reasons: An abundance of human thermal indices, a variety of interpretations of bio-meteorological terms, an array of procedures for data collection and a lack of agreed methods in determining thermal comfort ranges and index modifications. This study aims to review strategies and methods for human bio-meteorological research and to examine their suitability for thermal perception assessment. From 2001 to 2021, 254 case studies assessed human thermal perception by investigating in-situ thermal conditions versus subjective thermal perception, relying on protocols such as ASHRAE Standard 55 and EN ISO 10551 that were originally developed for indoor environments. Fifty-four cases determined different ranges for thermal comfort. Although 43 studies tried to modify indices to various climatic zones, only 13 studies modified the nine PET physiological stress categories and 4 studies modified the ten UTCI stress categories). Thus, comparisons between the studies' results become complicated. Our review points to three main reasons for the complexity: first, the 7-point TSV scales, does not always fit the scales of the applied thermal index; second, measurement procedures do not always represent the local climate conditions; third, certain methods for modifying thermal index scale thresholds are not capable of modifying the entire index scale. On the basis of our findings, we suggest a framework for bio-meteorological research, with attention to measurement procedure, appropriate questionnaire design, careful data control and suitable methods to enable modification of thermal indices. This study recommends applying systematic and objective statistical methods like linear regression and discriminant analysis in order to successfully modify the entire index scale.