Description
Description
In socrates_interface.f90 we have an option for using the Frierson-like sw profile:
https://github.com/sit23/Isca/blob/master/src/atmos_param/socrates/interface/socrates_interface.F90#L1091
In this option the coszen value is set to:
coszen = 0.25 * (1.0 + del_sol * p2 + del_sw * sin(rad_lat(:,:))).
This gives the same global-mean toa shortwave as any other reasonable profile (e.g. seasonal or perpetual equinox). But the maximum value of coszen this gives (if del_sol=1.4 and del_sw=0.0, which are the defaults) is 0.3375. This doesn't make much of a difference in the grey radiation code, but it matters significantly for Socrates. In fact, the tropical surface temperatures in a simple aquaplanet run with perpetual equinox vs frierson-sw are such that the tropical temperatures are significantly lower in the frierson case, even though the zonal-mean incoming sw is higher in the frierson case.
One proposal would be to change the frierson-sw profile in Socrates to include a diurnal cycle. This would allow coszen to change appropriately. But it's not entirely obvious what this ought to be. So for now let it be noted that the frierson_solar_rad option in Socrates should be used with caution.
An example of the comparison of the incoming sw schemes is given in the attached PDF. Frierson aquaplanets with grey radiation are shown with perpetual equinox vs frierson sw, and the same for Socrates. As can be seen, the grey radiation versions behave as expected - warmer tropics and cooler midlats with the frierson sw. But with Socrates this is not seen.
Activity