The base period is the time period used when estimating the needed adjustments of weather dependencies. If “Auxiliary – A” is used (A stands for Aiolos), this period is centered around the forecast start date one year before the forecasts starts and if “Auxiliary – C” is selected then the period will be immediately before the date for the last received outcome for the forecast series. The user can choose to let the reference period and base period coincide. The number of days during this period is specified under ”Length (d)”
The main predictor is the outcome values for the auxiliary series multiplied with the hourly day type quotients. Into the regression model, also 10 weather predictors are included, expressed as deviations during the base period from the hourly weather means already calculated. The method used for estimating the adjustments of weather dependencies is called stepwise multiple regression. In a step by step process it examines if any predictor should be included in the model or if any predictor already included should be excluded. More of this is explained in the model descriptions for Aiolos and Cilia. If the auxiliary and the forecast series do have similar weather sensitivity, this will be reflected in the result as low absolute values for the coefficients estimated, if any weather predictors included at all. Since the resulting equation will be a linear combination of the hourly day type quotients and the deviations from the day type weather, the regression also allows for a constant and a coefficient for the hourly quotient term.