Thermal loads on a building environmental system are greatly influenced by a number of weather variables. Dry-bulb temperature (dbt), net solar radiation intensity, dew point temperature (dpt), atmospheric pressure and wind velocity are some of these variables. To obtain a realistic estimate of the loads on a building the coincident patterns of dbt, dpt and solar radiation especially need to be considered. The set of algorithms compiled recently by ASHRAE and proposed for computer calculation of thermal loads and energy consumption estimates requires the use of coincident hourly weather data. It appears that the commonly used forms of weather data such as outdoor design temperatures, degree days, temperature level tabulations in cumulative hrs/season will not be adequate in fulfilling the newly created need for coincident data over extended periods of time.

One alternative is to use hourly weather data available from Environmental\Science Services Administration (ESSA) recorded on magnetic tape, cards or in printed form. The problems associated with these records are the extensive editing and decoding necessary to convert them into usable form. The printed form, in addition, requires to be put into a computer readable format. Perhaps the more imporant problem is the uncertainty inherent in selecting a representative period from the available data records, which extend over many years. A rigorous statistical analysis on weather variables that results in a mathematical representation of their coincident behaviour would provide an acceptable solution to these problems. The study repdrted herein is concerned with the feasibility of such an approach and its effects on thermal load computation methods.