How to change surface albedo correctly

[minimax]

Dear colleagues,

I need to run a simulation (F2000climo) with different albedo in a specific region.

As I understand from posts it this forum, to do this I need to change fsurdat file, in particular variable "SOIL_COLOR" for all months (including winter months).

First question - is it necessary to change also other variables in addition to "SOIL COLOUR" like LAI (variable "MONTHLY_LAI", to set LAI value for all types of plants to zero to exclude their influence) or something else to do?

Second question - I need to do experiments with a 10% reduction in albedo compared to the control experiment (F200climo). As I understand, for example, I have soil colour 10 associated with dry (vis 0.25 and nir 0.39) and saturated (vis 0.14 and nir 0.28). The new values will be
Dry VIS: 0.20
Dry NIR: 0.312
Saturated VIS: 0.112
Saturated NIR: 0.224
which roughly corresponds to soil colour 14 (2.3. Surface Albedos — ctsm release-clm5.0 documentation). Is this correct?

Thank you very much!

 

[slevis]

The model simulates the albedo using the soil color and soil moisture info, so you will unlikely get an exact 10% reduction in any given time step, but you can attempt your suggested change and see if you are satisfied with the result.

The answer to your first question depends on whether you are interested in reducing the soil albedo or the albedo seen by the atmosphere model, which includes vegetation growing above the soil, and possibly lakes, urban, and other landunits in the same grid cell.

 

[minimax]

The model simulates the albedo using the soil color and soil moisture info, so you will unlikely get an exact 10% reduction in any given time step, but you can attempt your suggested change and see if you are satisfied with the result.

The answer to your first question depends on whether you are interested in reducing the soil albedo or the albedo seen by the atmosphere model, which includes vegetation growing above the soil, and possibly lakes, urban, and other landunits in the same grid cell.

Dear Slevis,

thank you for the detailed and valued infromation!
Yes, in my case I need to change the albedo seen by the atmosphere model, the area where I need to change the albedo covers green grass in warm seasons and snow cover in winter. What is the best way to solve this in my case?
I think to run experiments separately for cold and warm seasons. At the moment I don't know how to run simulations for the differents seasons in CESM.

 

[slevis]

Ok, so for now you should focus on learning how to run the CESM.

My suggestion about reducing the albedo may be very difficult if you are not an experienced fortran programmer, and we do not offer support for that:
- You could make a control simulation with albedo values that you hardwire in the code.
- Then you could run an experiment simulation with 10%-reduced albedo hardwired in the code.
Because otherwise you need to figure out how to reduce the albedo of grasses and snow and soil, and you are still unlikely to get an exact 10% reduction.

 

[minimax]

Ok, so for now you should focus on learning how to run the CESM.

My suggestion about reducing the albedo may be very difficult if you are not an experienced fortran programmer, and we do not offer support for that:
- You could make a control simulation with albedo values that you hardwire in the code.
- Then you could run an experiment simulation with 10%-reduced albedo hardwired in the code.
Because otherwise you need to figure out how to reduce the albedo of grasses and snow and soil, and you are still unlikely to get an exact 10% reduction.

Dear Slevis,

thank you very much for your detailed reply!

 

[minimax]

Dear Slavis,
one more question, I probably gave the wrong experiment (F2000climo) information and it influenced your answer) In general it does not matter which experiment. Is there an experiment with "prescribed CLM" where I can change only CLM characteristics? or your previous answer still valid for this case.My aim is to investigate the atmospheric response to albedo change.

 

[slevis]

My answer is the same if you run land-only or coupled.