A novel method for local clothing insulation prediction to support sustainable building and urban design.

Clothing is crucial in thermal comfort evaluation, affecting heat exchange between the body and the environment. Assuming even clothing insulation across body segments can lead to inaccurate predictions, affecting building and urban design strategies. This study processed insulation data from 240 clothing ensembles to derive local insulation values. Regression models were developed to relate local and overall insulation values, allowing fast estimation of local insulation distribution for any given overall value. To validate the effectiveness of local clothing insulation values predicted by the proposed methods, measured real-time skin temperatures were collected from field experiments and compared with predicted values. Results demonstrated a significant accuracy improvement in the predicted local skin temperature from thermoregulation simulation combined with the proposed local clothing insulation estimation methods. The Jaccard Similarity Coefficient (JSC) increased by an average of 0.21, with body segments like the head, neck, shoulders, back, and arms showing nearly 0.4 or more improvement. These local insulation determination functions, used with a multi-nodal thermoregulation model, offer a simpler and more effective way to enhance thermal comfort assessment accuracy. By providing more precise local insulation values, these functions can help optimize building designs and urban planning strategies, leading to better thermal comfort for occupants.