A simple methodology is presented to predict local temperature and humidity. Such prediction is essential to forecast a building's thermal load. The site-specific prediction models have been tested with measured weather data at two different sites in Chicago, Illinois. The proposed models capture the influence of the localized weather by using the measured temperature and humidity at the site. The predicted models have demonstrated promising results. The models do not require any historical weather data, are easy to implement, and can be integrated in a building energy management system.

Units: I-P