aerosol scaling factors question

[dengxueliang9989@yahoo_com_cn]

Dear Cam3 Users:
i am freshman,now i have some difficult to undeestand the namelist terms about aerosol scaling factors ,for instance: CARSCL_RF , CARSCL, BGSCL_RF, DUSTSCL_RF,DUSTSCL,VOLCSCL,VOLCSCL_RF,SSLTSCL,SSLTSCL_RF and so
on,some is 1 in default and some is 0 in in default.
now i do not konw what they mean,and according to what to amend them.thanks for your help!

 

[dengxueliang9989@yahoo_com_cn]

can you tell me what 1 means and what 0 means? thanks

 

[bundy]

The aerosols implemented in CAM are prescribed from an offline climatology. This means that CAM reads an AerosolMass dataset, interpolates the data to the current model time, and uses those values of aerosol mass to do the radiation calculation.

The masses are multiplied by the scaling factors. The defaults are all = 1. If you wanted to run with 30% of the carbon, you do so by setting the namelist parameter CARSCL = 0.3 . You can check that this is working by looking at the carbon optical depth output field, CAROD_v, which should be reduced by the same scaling factor.

The scaling factors that end with _RF are in the case that you want to do a passive radiative forcing calculation. The usual radiation calculation is active: the heating rates calculated by the radiation are used to influence the model, changing, for instance, the temperature field. The passive calculation is different: the radiative forcing is calculated, written out (in fields like FSNS_RF, FSNT_RF, QRS_RF) and then thrown away: the passive calculation does not affect the model results. You can turn on the passive calculation with the namelist variable radforce = .true.

The scales for the passive calculation (e.g. CARSCL_RF) are by default = 0. This is so that, if you turn on radforce and don't change any scalings, the difference between the active and the passive calculations is the total aerosol radiative forcing.

 

[dengxueliang9989@yahoo_com_cn]

"The scales for the passive calculation (e.g. CARSCL_RF) are by default = 0. This is so that, if you turn on radforce and don't change any scalings, the difference between the active and the passive calculations is the total aerosol radiative forcing.“

sorry，i don't understand the last segment.Alought you turn on radforce,if you don't change any scalings including _RF ，the _RF by default also are 0,so the passive calculation also can't be calculated.
So i can't know how to get the total aerosol radiative forcing.


email:dengxueliang9989@yahoo.com.cn

 

[dengxueliang9989@yahoo_com_cn]

i follow your advice to add the namelist file with "radforce=.TRUE.",but i find that the output ''FSNS'',''FSNT'' and ''QRS''(i don't find' 'FSNS'_RF',''FSNT'_RF' and ''QRS_RF'' in userguide) is the same as the output in 'radforce=.FALSE.'
so this make me very puzzled.because you say that "This is so that, if you turn on radforce and don't change any scalings, the difference between the active and the passive calculations is the total aerosol radiative forcing",if following your method, i should see different in the namelist between the two different run.
can you give some suggest!
tks

 

[rneale]

You need to include the RADFORCE output variables (_RF variables e.g. FSNT_RF) in the namelist history file output as they are not included by default as with the the non _RF (eg FSNT) variables.
So to get the radiative forcing of a particular aerosol you only need to run the model once with RADFORCE set to .true. Then set the '_rf' scaling factor of the aerosol you want to examine eg sulphate to 1 so sulscl_rf=1.0. If you then run the model including the output as above then if you are in interested in net short-wave at the top of the model, FSNT will tell you what the value is including all the default aerosols and FSNT_RF will tell you what the value is with just sulphate aerosols included. The key thing to note is when RADFORCE is set to true it performs a second radiation calculation at each timestep with or without certain aerosols switched on. It is not use to integrate the model forward.

 

[dengxueliang9989@yahoo_com_cn]

dear rneale:
thanks for your help.waht you explain is very clear.i also have another question. does the passive caculation only influence the output *_rf ?and other outputs are controled by the default set(active caculation)?can the active and passive caculation work at the same time in a run?i don't understand.
thank you!

 

[dengxueliang9989@yahoo_com_cn]

dear richard:
thanks for your help.waht you explain is very clear.i also have another question. does the passive caculation only influence the output *_rf ?and other outputs are controled by the default set(active caculation)?can the active and passive caculation work at the same time in a run?i don't understand.
thank you!

 

[dengxueliang9989@yahoo_com_cn]

if i want to research the radiative forcing of all species aerosol , does the namelist 'radforce=.true.' have not to be set?i only change the namelist '*scl'=0 and =1 to separately run the model,then differernce of FSNT between two runs is equal to all aerosol radiactive forcing.Is this method correct?
thanks

 

[bundy]

Setting the namelist variable "radforce = .true." instructs the model to do two radiation calculations:

1) the active calculation
- is done regardless of 'radforce' setting
- output is written out in FSNT, FSNS, QRS
- results influence the model state (this is why it is called 'active')
- uses aerosol scales carscl, sulscl,... = 1 by default, so this calculation *includes* the effect of the aerosols by default.

2) the passive calculation
-is only done when radforce = true
-output is written out in FSNT_RF, FSNS_RF, QRS_RF
-results *do not* influence the model state (this is why it is called 'passive')
-uses aerosol scales carscl_rf, sulscl_rf,... = 0 by default, so this calculation *does not include* the effect of the aerosols by default.

For a single model run, with radforce on and using the defaults aerosol scales, the difference between the active and passive output fields gives the total aerosol forcing. For example, FSNT - FSNT_RF = net solar flux at the top of atmopshere due to aerosols.

If you do two runs with the active radiation difference:
1) all aerosol scales = 1
2) all aerosol scales = 0
the difference between the radiation fields is not just the aerosol forcing. This is because the active radiation calculation influences the model state (e.g. the heating rate QRS is used to adjust the model temperature) and in the two runs, this influence will be different. This is a way to calculate the aerosol *feedback* effect on the atmosphere, which is interesting also, but it is not the forcing. To calculate the forcing, use radforce (as explained above)

Some things to note:
1) Setting radforce = .true. will turn on the passive calculation, which does not influence the model, so FSNT etc should not change. Only FNST_RF will change
2) To test your set-up and understanding, you can do a 2-timestep run where you make the passive calculation identical to the active calculation. (by setting all {aer}scl_rf = 1). If FSNT_RF is not = FSNT, that is a sign something is wrong. If it is equal, then set the scales you want = 0, and make sure the results make sense.

By the way, which version are you running?

 

[dengxueliang9989@yahoo_com_cn]

thank you very much for your explanation!
i follow your advice to do a 2-timestep run (by setting all {aer}scl_rf = 1). the result is that FSNT_RF is = FSNT.i am very happy.with your help i think i can use the namelist to run aerosol effect of climate.
my plan is to compare aerosol optical thickness in model output with MODIS data,and then compute the aerosol radiation forcing .that seems too simple,can you give me some advice for my plan.
thank you！