prob_meteogram
prob_meteogram copied to clipboard
uncertainties for clouds
I see another issue with the way we represent clouds: Say after a couple of days the different ensemble members are actually quite close to a uniform random distribution of cloud covers between 0 and 1 for each cloud type, then the mean will be 0.5 for each. Do you think we should change the way how the clouds are plotted once the ensemble spread exceeds a certain threshold?