LAUSR

Abstract Most of thermal indices were developed based on the European/North American subjects and climates, while they should be investigated in other regions as well. This study aims to improve the applicability of nine thermal indices in Tehran, to be used in heat-health warning systems. 1008 interviews were conducted along with field measurements, and modified scales of thermal indices were determined using three different techniques. The results revealed low percentages of correct predictions for original scales of thermal indices despite having high correlation coefficient with thermal sensation votes–TSV, while modified scales of Universal Thermal Climate Index–UTCI and Physiologically Equivalent Temperature–PET correlated better with TSV and had higher correct predictions percentages than the original scales, especially in the ‘neutral’ class. Also the modified scales determined by ‘Probit analysis’ and ‘PD diagram's fitted curve’ techniques correlated better with TSV than those acquired by ‘linear regression’. Comparing the modified PET and UTCI scales with studies conducted in similar climates revealed differences between the ‘cool’ and ‘cold’ categories, probably caused by cultural issues. The results of this study can help urban and landscape designers to design more comfortable outdoor spaces.