Sulfur in CAM

CAM can calculate the radiative forcing from sulfur in two modes: "prescribed" or "prognostic". In prescribed mode, CAM reads in a 3D monthly mean sulfate aerosol dataset to estimate the concentration of sulfate aerosols for radiative transfer calculations. This is the CAM default and occurs automatically in every simulation without any user input. Prognostic mode simulates the entire sulfur cycle, including emissions, transport, chemistry, deposition, etc. This page will provide a brief overview of prognostic Mode and how to conduct simulations in this mode.

Overview of the Prognostic Sulfur Cycle in CAM

CAM predicts the mixing ratios of DMS, SO2, SO4, and H2O2. O3 is a prescribed field. The model conducts Emissions, Transport, Chemistry, and Deposition. For further details, see Barth et. al, JGR, 2000.

Emissions

CAM utilizes emissions datasets for DMS and SO2, the two key sulfate precursors, and calculates dynamic mixing ratios of these two species, as well as SO4 itself. CAM also utilizes an oxidizers dataset which is used to calculate the reactions of SO2 and DMS, but the concentrations of these oxidants is fixed (not prognostic). These emissions datasets are in netcdf format (.nc extension) and there is one file for each species for each model resolution. The emissions files originate from the latest GEIA inventory and mass mixing ratios are calculated for each grid box. Make sure that you are using the most up to date file. There are some files floating around with "noncon" in the name; these do not conserve mass and should not be used.

Transport

CAM's dynamical core calculates winds and the model advects each species with the winds. Because of their high solubility, sulfate aerosols are not convectively transported, however.

Chemistry

There are 6 gas-phase reactions and 3 aqueous reactions included as sulfur chemistry. DMS, SO2, and SO4 mass mixing ratios are actively traced, while all other chemical species (including OH, H2O2, etc) have assumed, constant concentration fields that are read in and are not an active part of the model.

Deposition

CAM incorporates both wet deposition and dry deposition.

Wet Deposition - Deposition rates are calculated separately for gases and aerosols. For gases, Henry's Law is used to determine the fraction that is contained in liquid water precipitation. (Gases are not deposited via snow). For aerosols, the model can undergo both in-cloud and below-cloud scavenging. 10% scavenging efficiency is assumed.

Dry Deposition - The dry deposition rate for SO2 is inversely proportional to the sum of the aerodynamic resistance, the resistance to transport across the atmospheric sublayer in contact with surface elements, and the surface resistance. The dry deposition velocity for SO4 is inversely proportional to the sum of the aerodynamic and sublayer resistances.

Relevant CAM files to be aware of

sulfur_intr - module to interface aerosol parameterizations with CAM

How to Run Prognostic Sulfur in CAM

By default, CAM calculates radiative forcing for sulfate aerosols in "prescribed" mode. These instructions will describe how to run a simulation in CAM in "prognostic" mode. "Prognostic" mode incorporates the entire sulfur cycle, while "prescribed" mode simply reads in a 3D monthly mean sulfate aerosol dataset to estimate the concentration of sulfate aerosols for radiative transfer calculations.

List of required actions by the user

Tracers

Before building CAM, instruct CAM to keep track of four additional tracers (DMS, SO2, SO4, H2O2):

setenv CAM_CONFIGURE_OPTS '-nadv 7'

In this example, 7 tracers are required based on CAM's default of 3, plus the sulfur cycle 4. If your simulations have more than 3 tracers, add 4 to this number and use the command above.

Emissions datasets

Add links to the namelist file to tell CAM where to look for the three emissions datasets. The default location is in CAM's data/atm/cam2/scyc folder. The example below is for a 4x5 resolution run based on 2006 emissions datasets. Note: It is important to use the latest emissions datasets. There may be some 2005 datasets floating around with "noncon" in the name; these do not conserve mass and should not be used.

bndtvsox	= '/Volumes/Data/Models/cam/trunk/data/atm/cam2/scyc/SOx_emissions_4x5_L2_1990-2095_c061114.nc'
bndtvdms	= '/Volumes/Data/Models/cam/trunk/data/atm/cam2/scyc/DMS_emissions_4x5_clim_c061114.nc'
bndtvoxid	= '/Volumes/Data/Models/cam/trunk/data/atm/cam2/scyc/oxid_4x5_L26_clim_c061114.nc'

Prognostic/Prescribed mode switches

Add a line to the namelist file to switch prescribed mode to passive, and another line to activate the prognostic subroutines in "passive" or "direct" mode. All of the possible settings for each of the two modes are shown below. One of them must be set to "direct".

prescribed_sulfur = 'direct' [default], 'passive'
prognostic_sulfur = 'off' [default], 'passive', 'direct'

Following is a description of each of the three prognostic modes:

'off' means that prognostic aerosol subroutines are turned off (radiative forcing is calculated from prescribed aerosols).
'passive' means that the prognostic cycle calculates mass mixing ratios and optical depths, but the calculations are not used to calculate radiative forcing (radiative forcing is calculated from prescribed aerosols). This mode is useful if one wishes to simulate the sulfur cycle without having the prognostic aerosols directly affecting the radiative forcing, which could cause positive feedbacks.
'direct' means that the prognostic cycle calculates mass mixing ratios and optical depths, and these values are used to calculate the radiative forcing. This mode fully simulates the complete sulfur cycle and its link to radiative forcing. This is the most comprehensive mode; but there is some risk that an error somewhere can cause inaccurate climate feedbacks.

Simulation year designation

By default, CAM starts at year 1. However, in order to run prognostic sulfur, CAM must run within the dates of the emissions datasets (1990-2095). Enter the following line in the Namelist file to change the starting date to January 1, 2005, for example:

start_ymd=20050101

Radiative Forcing Flag

Add the following line to the CAM Namelist file to compute the radiative forcing from aerosols.

radforce = .true.

Passive radiation calculation (not for forcing)

When running a prognostic simulation in "passive" mode, enter the following line in the namelist file to ensure that CAM conducts the appropriate radiative calculations. Even though these calculations will not affect the radiative forcing, the fields will still be calculated for your analysis. Think of it as an off-line simulation.

sulscl_rf = 1

This scaling factor can be changed. A value of zero means that the calculation is not done. A value other than 1 will scale the calculations accordingly.

Active radiation calculation (for forcing)

In the Namelist file, set the sulscl factor to 1 to tell CAM to calculate the radiative forcing from sulfur. (This should be its default value anyway).

sulscl = 1

This scaling factor can be changed. A value of zero means that the calculation is not done. A value other than 1 will scale the calculations accordingly.

Write relevant fields to disk

By default, CAM does not write some important aerosol fields to disk. Include these by adding the following line to the Namelist file. To ensure that all fields write time-averaged values, add an ":A" after the fields which have instantaneous values as their default. A list of all the fields and their default averaging flag (AF) are provided in the CAM User's Manual. To conserve disk space, use the empty tapes command first, then include all desired fields in the fincl1 command as shown below.

empty_htapes = .TRUE.

fincl1 	= "DMS:A", "SO2:A", "SO4:A", "H2O2:A","AERASM_v:A", "AERFWD_v:A", "AEROD_v:A", "AERSSA_v:A", "ALDIF", "ALDIR", "ASDIF",
"ASDIR", "BGOD_v:A", "CLDTOT", "CLOUD", "FSDS", "FSDSC", "FSNS", "FSNSC", "FSNT_RF", "FSNS_RF", "FU:A", "FV:A", "MSO4", "MVOLC:A", 
"PRECC", "PRECCav", "PRECL", "PRECT", "QRS_RF", "SRFRAD", "SULFANT","SULFBIO", "SULFMMR", "SULOD_v:A", "T", "TS", "U", "V"

Add any other desired fields to the list above for your specific simulation.

Examples

Example 1: 4x5 Prognostic simulation with interactive climate

In this example, a prognostic simulation will be run at 4x5 degree resolution in CAM 3.1, where the radiative forcing is interactively coupled to the aerosol calculations (e.g. climate/aerosol feedbacks are possible). The most up-to-date emissions datasets are used (2006 in this case), a simulation will be run for 20 years, starting at 2005, data written every 24 hours, and 60-days worth of data lumped into each .nc file.

Enter the following commands at a terminal window prompt in the CAM/trunk folder.

setenv CAM_RES 4x5
setenv CAM_CONFIGURE_OPTS '-nadv 7'
scripts/make-cam.csh   Y   Y   Y

Then, open the Namelist file in the Build directory with a text editor and make the following additions:

bndtvsox	= '/Volumes/Data/Models/cam/trunk/data/atm/cam2/scyc/SOx_emissions_4x5_L2_1990-2095_c061114.nc'
bndtvdms	= '/Volumes/Data/Models/cam/trunk/data/atm/cam2/scyc/DMS_emissions_4x5_clim_c061114.nc'
bndtvoxid	= '/Volumes/Data/Models/cam/trunk/data/atm/cam2/scyc/oxid_4x5_L26_clim_c061114.nc'

start_ymd      = 20050101
stop_ymd       = 20201231
mfilt(1) 	= 60
nhtfrq(1) 	= -24

prognostic_sulfur = 'direct'
prescribed_sulfur = 'passive'
radforce       = .TRUE.
sulscl_rf      = 0
sulscl         = 1

empty_htapes = .TRUE.
fincl1 	= "DMS:A", "SO2:A", "SO4:A", "H2O2:A","AERASM_v:A", "AERFWD_v:A", "AEROD_v:A", "AERSSA_v:A", "ALDIF", "ALDIR", "ASDIF",
"ASDIR", "BGOD_v:A", "CLDTOT", "CLOUD", "FSDS", "FSDSC", "FSNS", "FSNSC", "FSNT_RF", "FSNS_RF", "FU:A", "FV:A", "MSO4", "MVOLC:A", 
"PRECC", "PRECCav", "PRECL", "PRECT", "QRS_RF", "SRFRAD", "SULFANT","SULFBIO", "SULFMMR", "SULOD_v:A", "T", "TS", "U", "V"

Finally, enter the following commands at a terminal window prompt:

setenv CAM_RUN example_1
scripts/run-cam.csh

Example 2: 4x5 Prognostic simulation with passive climate

In this example, a prognostic simulation will be run at 4x5 degree resolution in CAM 3.1, where the radiative forcing is NOT coupled to the prognostic aerosol calculations (e.g. prescribed aerosols are used to calculate climate forcing). The most up-to-date emissions datasets are used (2006 in this case), a simulation will be run for 20 years, starting at 2005, data written every 24 hours, and 60-days worth of data lumped into each .nc file.