CAM[4,5] with the Sun zenith angle fixed, no seasonal cycle

[gus@ldeo_columbia_edu]

Hi CAM/CESM experts

I want to run a CAM[4,5] experiment with no seasonal cycle,
the sun zenith angle fixed, i.e., no seasonal cycle.
There are two variants: one with daily cycle,
another without daily cycle.

Is this possible in CAM[4,5]?
How can it be implemented in the namelists?

I thought of using the namelist below for the the first variant,
then adding
solar_data_tod=0
for the second variant:

&solar_inparm
solar_const = 1365.0
solar_htng_spctrl_scl = .false.
solar_data_type=FIXED,
solar_data_ymd=20000701
/

However, the code seems to ignore all namelist parameters when
'solar_const' is set to a value.

Is there a workaround?

Thank you,
Gus Correa

 

[eaton]

Early versions of CAM (and the CCM model that preceeded it) had support for
the "perpetual mode" that you're asking for. This mode is no longer
supported, but I believe it could be achieved (at least in CAM) with a
reasonable amount of work. The main issue is that the interpolation of
boundary datasets needs to respect the perpetual date rather than the
simulation date.

The perpetual date is supported in CAM's time manager code via the namelist
variables perpetual and perpetual_ymd. Essentially this forces the time
manager's get_curr_calday function to always return the calendar day
associated with the perpetual date while accounting for the diurnal cycle.
You'll need to go into the get_curr_calday function and modify code to
remove the diurnal cycle.

The solar zenith angle is set by calls to subroutine zenith which has
"calday" as an argument. In all calls to zenith the calday argument is
obtained from the get_curr_calday function. All this is independent of how
the solar constant is set.

The problems with perpetual mode revolve around the fact that the current
boundary dataset interpolation schemes don't look to see whether the
perpetual mode is active, and hence don't use the perpetual date for
interpolation. However, there are workarounds for this. All (or most?)
the boundary interpolations in CAM have namelist options that allow fixing
the date to which the boundary data will be interpolated. For example,
look at the namelist variables that are controlling the ozone and aerosol
datasets and you'll find these options.

Finally there are the surface components to deal with. This may be the
biggest challenge. Back when CAM/CCM supported the perpetual mode it was
released as a standalone model with its own versions of the surface
components which recognized the perpetual mode. Now it is only released as
part of CESM and I'm not sure whether the surface components have a
perpetual functionality or not. You may need to explore that with the
maintainers of those components. Of course if you're interested in a
special mode like aquaplanet then this will not be a problem.