Utilizing Reanalysis Data in MOS or Perfect Prog

Statistical postprocessing of numerical weather prediction model forecasts provides a means of refining the latter. MOS and Perfect Prog are two popular approaches. Here, an alternative method (called RAN) is examined that combines the two, while simultaneously utilizing the information in reanalysis data. The three methods are examined from a purely formal/mathematical point of view within a regression framework. The results suggest that whereas MOS is expected to outperform Perfect Prog and RAN in terms of mean squared error, bias, and error variance, the RAN approach is expected to yield more certain and bias-free forecasts.