Setting CAM 3.0 to not pure aqua planet

[san3@meteo_itb_ac_id]

I have problem to simulate aqua planet simulation, which it's not pure aqua planet in CAM 3.0. I want to let a few islands or continent in aqua planet. So, what should i do to make the simulation can be run? Where should I modify the model's component?
Anybody can help me?
Thank you.
san3

 

[olson]

If you want these continents in the aquaplanet model for their topographical
effects only, you need to go into cam1/models/atm/cam/src/dynamics/fv/inidat.F90
(or .../eul/inidat.F90 for Eulerian dycore) and delete code that sets
"phis_tmp" to 0. Then create a "PHIS" field that the model can read in that
has the topography of the land features you want.

If, however, you want land processes to interact with the model, I would
suggest doing the inverse of what you're proposing: Run the full model rather
than aquaplanet, but replacing the standard climatological SST's with
aqua-planet SST's as defined in
cam1/models/atm/cam/src/ocnsice/dom/sst_data.F90 (using the formula
for "sst_option=1").
Furthermore, you would need to modify various other external datasets such as
land datasets to be consistent with your land definitions.

 

[san3@meteo_itb_ac_id]

Hi, Olson
Thanks for your reply.

But I have more question again. What parameter exactly contains in phis_tmp variable and what the unit of it? When I look the variable, the maximum value is 50.000 and the minimum value is 2000.
If I changes phis_tmp, should I changes the other variable with zeroed in aqua planet run, such as SGH, SGH30, LANDM, LANDFRAC, and also what parameter contains in those variable? My guess LANDM is landmask and LANDFRAC is landfraction. Am I correct? But I don't know about the other variables.

Regards
San3

 

[olson]

phis_tmp is the surface "geopotential" defined as height of the land surface above sea-level times
"g". phis = Zs*9.80616 the units are m**2/s**2

If all you're interested in is the topographical effects on aquaplanet then the other fields you
mention would all be 0. We call this the "water mountain" effect in aqua-planet.

If, as I stated before, you really are interested in defining land features such as vegetation, desert,
soil types, etc., then you would want to run full model, not aqua-planet. "landm" is a 0.-1. land halo
ramp around all land. Plot the full model landm and you will see what I mean. You would need to define
your own ramp for your land features. sgh and sgh30 are the standard deviation of sub-gridscale land
orography. I presume you could just set those to 0. Yes, landfrac is land fraction and you would need to
define that as well. Plus define all the land features I mentioned earlier (vegetation, soil, etc.). This is
much more complicated than if you're just interested in "water mountains".