Diffuse radiation in multilayer model

There was a mistake in the calculation of the diffuse radiation in the new VOM, I used the method of Roderick for hourly values, but this is only valid for daily values. I corrected this now based on Spitters et al. (1986), who basically assume that the ratios Idiff/Iglob remain constant over the day. This is not the case for the data in the multilayer model:

@stanislaus.schymanski, do you know how the diffuse radiation was determined in the multilayer model? I just have it as a time series, and don't know if it was derived, or somehow observed.