Why does there appear to be a diurnal variation in snowcover?
It has been a well-known problem for several years or more, but less so in RR than in current ops Eta, and much less so than before the major Eta/Noah land-surface upgrade in July 2001. Specifically, going back to before July 2001, about HALF of the magnitude of the fast snowmelt bias in the Eta was solved with the major Eta/Noah land-surface upgrade in July 2001. Of the remaining bias, about half of that was solved in the RR snow albedo and snow fraction upgrades put into the RR shortly before the RR production phase. The remaining fast snowmelt bias in the RR is due I feel to the high bias in solar insolation in the Eta and RR. The Eta/RR has a high solar insolation bias because the Eta-RR solar radiation fails to attenuate enough radiation through cloud cover — a problem quite substantial in the high cloudiness regime of winter season. So, the Eta/RR has a widespread tendency to melt off snow cover too fast during the day (not always, depends on air temperature and solar insolation
