Artificial neural network models as a useful tool to forecast human thermal comfort using microclimatic and bioclimatic data in the great Athens area (Greece).

The present study deals with the development and application of Artificial Neural Network (ANN) models as a tool for the evaluation of human thermal comfort conditions in the urban environment. ANNs are applied to forecast for three consecutive days during the hot period of the year (May-September) the human thermal comfort conditions as well as the daily number of consecutive hours with high levels of thermal discomfort in the great area of Athens (Greece). Modeling was based on bioclimatic data calculated by two widely used biometereorogical indices (the Discomfort Index and the Cooling Power Index) and microclimatic data (air temperature, relative humidity and wind speed) from 7 different meteorological stations for the period 2001-2005. Model performance showed that the risk of human discomfort conditions exceeding certain thresholds can be successfully forecasted by the ANN models. In addition, despite the limitations of the models, the results of the study demonstrated that ANNs, when adequately trained, could have a high applicability in the area of prevention human thermal discomfort levels in urban areas, based on a series of relatively limited number of bioclimatic data values calculated prior to the period of interest.