Hello everyone,
I am developing a surface spectral PAR forcing for my own model and I am using the ROMS spectral-light framework as a reference.
I have been looking at mod_eclight, ec_Fobar, and ana_specir.h, and I understand that:
ec_Fobar represents mean extraterrestrial spectral solar irradiance
ana_specir.h computes sea-surface spectral irradiance (SpecIr)
ROMS then uses this spectral information in the biology/light calculations
My input data are from ERA5. Since ERA5 surface shortwave radiation already includes the effects of the atmosphere, I assume that the main atmospheric corrections are already accounted for in the broadband forcing.
My question is:
For my own model, can I use normalized ec_Fobar as spectral weights to split ERA5 broadband shortwave radiation or PAR into wavelength bands?
For example, would it be reasonable to do:
wi = ec_Fobar(i) / ∑iec_Fobar(i)
and then compute: PARi = PARtotal *wi
Or would this still be physically inconsistent, because ec_Fobar is a top-of-atmosphere spectrum and not a sea-surface spectral distribution?
Since ERA5 already represents surface radiation after atmospheric effects, I am wondering whether using ec_Fobar only as a normalized spectral shape would be acceptable for constructing surface spectral PAR for my model, or whether a ROMS-like ana_specir approach is still necessary.
My goal is to generate surface spectral PAR from ERA5 broadband radiation for use in my own model.
Thank you very much.