Can the forecast N+1 days ahead be more certain than the forecast N days ahead?

Question

The Swedish meteorological agency SMHI on its website (NB: freezes Firefox on Linux for me, likely due to plugins, works in Chrome) indicates how certain the forecast for a particular day is. Looking at today (Wednesdays) forecast for the coming week for Kvikkjokk, they estimate that the forecast for Friday is ganska säker (somewhat certain), for Saturday is osäker (uncertain), and for Sunday is säker (certain); they estimate that the forecast for Sunday is more certain than the forecast for either Friday or Saturday. I suppose that what they do is look at the forecast ensemble and estimate certainty based on agreement between different runs from the ensemble. But as the state of the atmosphere on Sunday should depend on the state on Friday and Saturday, is there any way how it can make sense that the forecast for Sunday is more certain than the one for Friday or Saturday?

Answer 1

Typically, days with conditions which might produce local instabilities would produce a less certain forecast than a day with a stabilize pattern even if the more stable day was further in the future. For instance, tomorrow may have the potential for localized disturbances some maybe there will be thunderstorms, maybe they will not form. But two days later, a large high pressure zone will almost certainly clear everything out or a band of rain will definitely reach the area.