May be superseeded by other more specific issues, but capturing here that when we are able to aggregate over a number of case studies, there are a variety of ways to do this. This includes as a function of forecast lead-time, time of day (diurnal cycle), and as a longer timeseries (validity time - though there might be some overlap if a forecast runs every 24 hours but runs out to 48h).
Also capturing thoughts around quantifying uncertainty/spread (done in the RES through confidence bars) - this is more of a nice to have than specific requirement for now.
What problem does your feature request solve?
May be superseeded by other more specific issues, but capturing here that when we are able to aggregate over a number of case studies, there are a variety of ways to do this. This includes as a function of forecast lead-time, time of day (diurnal cycle), and as a longer timeseries (validity time - though there might be some overlap if a forecast runs every 24 hours but runs out to 48h). Also capturing thoughts around quantifying uncertainty/spread (done in the RES through confidence bars) - this is more of a nice to have than specific requirement for now.