This dataset has been set up on a daily S grid even though it is pentadly, as in 5-daily from 1-Jan on a 365-day calendar, ie ignoring 29-Feb always in the 5-day step. This set up makes the objects 5 times bigger than needed, and forces the Data Library to seek for files that have never existed and will never. For better performance, consider the version set up on a pentadly daily grid.