Hourly weather data are needed for building energy calculations; however, many available data from previous years are 3-hr-based. Therefore, appropriate interpolation methods should be employed in order to fill the gaps of the data. In this paper, linear and cubic spline functions are used to interpolate the weather data of Kerman, Iran. Weather data have different behaviours in heating and cooling seasons. Therefore, this paper analyses the interpolation methods in heating and cooling seasons separately. In order to specify all the weather psychrometric characteristics, three independent parameters are required. First, temperature and pressure were taken as two of the parameters. Then, humidity ratio and relative humidity were tested and analysed as the third parameter. Furthermore, the methods of filling weather data gaps were examined in particular to their effects on calculating accurate energy consumption. The results showed that the best interpolation method for each variable depends on the criterion of accuracy. Each method gives the best accuracy for a specific case. However, for temperature and pressure data, cubic splines were more accurate than the linear ones in all cases. For other weather data, it is important to consider the purpose of filling the gaps. The appropriate criterion of accuracy is different according to whether the average or the extreme values of the data are required. For calculation of the energy consumption variables, one interpolation technique should be used for all weather elements; linear interpolation by using relative humidity as the third psychrometric variable yielded better results for this aim.
K E Y W O R D Scooling season, heating season, hourly weather data, linear interpolation, missing data, spline functions, temporal gaps