Multi-Method Integrated Approach to Assess Human Climate Comfort in Iran

Abstract

Understanding human thermal comfort is essential for assessing environmental conditions and their implications for well-being, particularly in the context of global climate change. This study examines the influence of 30 climatic and ecological factors, including temperature, humidity, atmospheric pressure, solar radiation, wind dynamics, and topographical characteristics, on human thermal comfort across Iran. A multidisciplinary approach was employed, integrating principal component analysis (PCA) for feature selection, multivariate regression (MR) for impact quantification, cluster analysis (CA) for climate classification, and spatial modeling (SMA) to assess regional disparities. Furthermore, machine learning models (MLMs) and artificial neural networks (ANNs) were utilized to capture complex, nonlinear relationships in climate–comfort interactions. Based on a comprehensive data set spanning 38 years (1984–2022), the findings reveal significant spatial variations in climate sensitivity. Weighted indices such as predicted mean vote (PMV), physiologically equivalent temperature (PET), and thermal discomfort index (TDI) enhance the precision of comfort assessments. The results indicate that northern Iran, particularly the western coastal region of the Caspian Sea, exhibits the most favorable climatic conditions, whereas arid and semi-arid areas experience heightened thermal stress. These insights advance biometeorological research by linking climate variability to human physiological responses and provide practical implications for urban planning, public health policies, and climate adaptation strategies. By integrating high-dimensional climate data with advanced computational techniques, this study highlights the necessity of adaptive measures to mitigate the impacts of climate change on human thermal comfort.