MADE3  3.0
Modal Aerosol Dynamics for Europe, adapted for global applications, 3rd generation
 All Classes Functions Variables Pages
Couplings and output

For an example setup see messy/nml/REF_2000_MADE3 (can be obtained from the MADE3 maintainer upon request).



The submodel AEROPT calculates aerosol optical depth (AOD), an overall (i.e., integrated over all modes) aerosol single scattering albedo, and an overall aerosol asymmetry parameter. These properties can subsequently be used as an input for the radiation submodel RAD , and they can be output for diagnostic purposes. For the required calculations, AEROPT uses pre-calculated lookup tables of the aerosol particle extinction cross section, single scattering albedo, and asymmetry parameter as a function of the Mie size parameter, $y$, and the particle complex refractive index, $\varepsilon$. The latter two are computed from the MADE3 aerosol properties. For each mode, $y$ is calculated as $y=\frac{\pi D_{\rm p}}{\lambda}$, with the median mode diameter $D_{\rm p}$ and the wavelength $\lambda$, and $\varepsilon$ is determined as the volume average over the refractive indices of the mode's components. These, again, are tabulated for a number of species as a function of wavelength.

The above-mentioned tables have to be supplied separately for the shortwave and for the longwave part of the spectrum. Their paths must be specified via the CTRL namelist in aeropt.nml. For MADE3, only one set of files is currently available. Hence, unless you create your own lookup tables, the appropriate line for the AEROPT CTRL namelist is:

read_lut_sets(n) = m, "$INPUTDIR_MESSY/../EVAL2.3/messy/raw/aeropt/", "$INPUTDIR_MESSY/../EVAL2.3/messy/raw/aeropt/",

The indices n and m have to be set in accordance with the rest of your aeropt.nml.

In order to make AEROPT aware of the MADE3 aerosol, a further entry is required in the CPL namelist in aeropt.nml. Use

read_aero_sets(l) = "MADE3" , t, "", "", "", t, f, "made3_gp", "wetradius", "dryradius", "", m, 0.55,,,,,,,,,,,,,,,,

here, set the index l in accordance with the rest of your aeropt.nml, use the index m that you defined in the CTRL namelist entry, and adapt the "0.55,,,,,,,,,,,,,,,", i.e., up to 16 wavelengths (in micrometers) at which AOD is calculated for diagnostic purposes, according to your needs.

Leave the rest of the settings as they are, unless you have read and understood the code where they are used.


The submodel CLOUD will be able to diagnose CCN from the MADE3 aerosol, and will eventually also be able to use MADE3 aerosol as an input for cloud formation. These features are under development and will be documented here as soon as they have been tested.


The submodel DDEP automatically uses the parameters wetradius, densaer, and sigma supplied by MADE3 (see * to compute dry deposition of aerosol particles (excl. sedimentation, for which SEDI is used). The only MADE3-specific user setting is in the CPL namelist in ddep.nml: add the full names of the tracers for which you want to output the dry deposition flux to the variable outflux (separated by ";"). A full name is made up of a base name, an underscore, and a subname, as explained in *


The submodel MECCA does not require any specific settings for MADE3 in mecca.nml, as the interaction happens via tracers, which are specified via the CPL namelist in made3.nml .

MADE3 has so far only been run (and evaluated) with a simplified chemistry mechanism. The code (and pdf file) for this "meccanism" can be created via xmecca, which can be run from within messy/util/xconfig. The selection of reactions is done via the batch file messy/mbm/caaba/mecca/batch/user_contributed/simple_MADE.bat (in xmecca, first choose no. 3 ("SUBDIRECTORY: ./user_contributed"), then no. 12). A replacement file for certain reactions is used by this batch file (messy/mbm/caaba/mecca/rpl/mim1-simple_MADE.rpl).


The submodel RAD will be able to compute the effect of the MADE3 aerosol on solar and terrestrial radiation. This feature is under development and will be documented here as soon as it has been tested.


The SCAV routines for cloud processing of aerosol particles differ for different aerosol submodels in some cases. Hence, the SCAV code contains a number of MADE3-specific parts. Nevertheless, a few parameters can be controlled via the namelist file scav.nml.

Specifically, these are:

  • frac_resnum in the CTRL namelist, the factor applied to the number of cloud residual aerosol particles upon cloud evaporation. This factor is applied to all the MADE3 residual modes (as, am, cs, cm; for mode naming see *, rather than only to the hydrophilic mode with the largest particles (default in SCAV). The value used in the evaluation of EMAC with MADE3 is
    frac_resnum = 0.1,
  • aermod_gp in the CPL namelist has to be set to
    aermod_gp = 'made3',
  • made3_params in the CPL namelist should be set to
    made3_params = 1,2,3,4,5,6,7,8,9,9,8,'cm',8,4,
    unless you make changes to the MADE3 and/or SCAV code. The first nine numbers are the indices of the MADE3 modes in the order ks, km, ki, as, am, ai, cs, cm, ci (for mode naming see *, followed by the total number of modes. These values should only be changed in case the MADE3 code is modified. Next, the index and subname of the mode for dummy tracers are supplied. SCAV may create such tracers to ensure its chemical mechanism works properly. These settings are mainly "cosmetic". The last but one parameter is currently not used, and the last one specifies how many residual modes shall be considered. This should only be changed if the MADE3-specific SCAV code is modified.
  • out_string and out_string_aer in the CPL namelist determine for which species and tracers SCAV creates wet deposition flux channel objects. For full diagnostic output, and in order to enable the (post-simulation) calculation of residence times for all species, they should be set to
    out_string = 'H2SO4_l,HSO4m_l;SO4mm_l;HNO3_l;NO3m_l;NH3_l;NH4p_l;HCl_l;Clm_l',
    out_string_aer = 'SO4_ks;SO4_km;SO4_ki;SO4_as;SO4_am;SO4_ai;SO4_cs;SO4_cm;SO4_ci;NH4_ks;NH4_km;NH4_ki;NH4_as;NH4_am;NH4_ai;NH4_cs;NH4_cm;NH4_ci;NO3_ks;NO3_km;NO3_ki;NO3_as;NO3_am;NO3_ai;NO3_cs;NO3_cm;NO3_ci;Na_ks;Na_km;Na_ki;Na_as;Na_am;Na_ai;Na_cs;Na_cm;Na_ci;Cl_ks;Cl_km;Cl_ki;Cl_as;Cl_am;Cl_ai;Cl_cs;Cl_cm;Cl_ci;POM_ks;POM_km;POM_ki;POM_as;POM_am;POM_ai;POM_cs;POM_cm;POM_ci;BC_km;BC_ki;BC_am;BC_ai;BC_cm;BC_ci;DU_am;DU_ai;DU_cm;DU_ci;N_ks;N_km;N_ki;N_as;N_am;N_ai;N_cs;N_cm;N_ci',
    (where the latter should actually be on one line)
The SCAV–MADE3 interplay has only been thoroughly tested with the SCAV settings in messy/nml/REF01_2000_MADE3/scav.nml. Hence, it cannot be assumed that other settings that work with GMXe, for instance, will also work with MADE3, even if they should not be submodel dependent.


The submodel SEDI automatically uses the parameters wetradius, densaer, and sigma supplied by MADE3 (see * to compute sedimentation of aerosol particles. It does not require any MADE3-specific settings in sedi.nml.


Tracer properties are set in messy_made3_si::made3_new_tracer using subroutines supplied by the submodel TRACER. The relevant settings for coupling to the base model and to other submodels are those for:

  • transport
    • I_ADVECT = ON for all aerosol tracers, except aerosol water, i.e., all aerosol tracers except aerosol water are advected
  • deposition
    • I_WETDEP, I_DRYDEP, I_SEDI are all set to ON, i.e., wet and dry deposition, as well as sedimentation, are switched on for all aerosol tracers
  • scavenging
    • I_SCAV = ON for all aerosol tracers, i.e., scavenging is switched on for all aerosol tracers
    • I_AEROSOL_SOL = ON for the soluble and mixed modes, OFF for the insoluble modes, i.e., the latter are considered hydrophobic, whereas the former are considered hydrophilic
    • I_AEROSOL_HETICE = ON for the mixed and insoluble modes, OFF for the soluble modes


Depending on the settings in your channel.nml, you can have the following output of MADE3 aerosol related quantities. For an example setup with typical output used for aerosol evaluation see messy/nml/REF01_2000_MADE3/channel.nml.

If you choose output of non-instantaneous values in channel.nml the following variable names will have a further extension, like "_ave" for example.


  • aot_[ls]w_BNN, gamma_sw_BNN, omega_sw_BNN contain AOD, asymmetry parameter, and single scattering albedo per radiation band (extension "BNN" for band no. NN); some of these can be used to couple the MADE3 aerosol to the radiation submodel RAD
  • aot_opt_<mode>_<wavelength>_<species> contain diagnostic AOD, where <mode> (either "MNN" or "TOT") corresponds to the MADE3 mode with index NN or the sum over the modes ("TOT"), <wavelength> is determined by what you request via the CPL namelist in aeropt.nml and <species> stands for one of the AEROPT particle component classes (or the sum over all, in which case <species> is "total"):
    • bc (black carbon)
    • du (mineral dust)
    • h2o (aerosol water)
    • oc (organic carbon)
    • ss (sea spray)
    • waso (other water soluble components)
    These variables contain the AOD per layer, i.e., for comparison to satellite measurements, for instance, the values of all layers have to be summed.


  • ddepflux_<tracer> contain the dry deposition fluxes of the <tracer>s (see * in units of molecules (or particles, for number tracers) per square meter per second
  • ddepfluxsum_<tracer> contain the time integrals of the above, converted to mass (i.e., [kg m-2]) in case of mass tracers


In this subsection <modename> always stands for the two-character combinations explained in *

  • wetrad_<modename> contain the modal median particle wet radii
  • dryrad_<modename> contain the modal median particle dry radii
  • densaer_<modename> contain the modal particle densities
  • rhhist_<modename> contain the modal deliquescence histories
  • sigma contains the MADE3 mode widths
  • burden_* and sink_* should contain data to determine the conversion lifetime of hydrophobic to hydrophilic black carbon (BC) and mineral dust (DU), respectively, but these have not been used so far, so they might not contain the expected data


  • <source>_<tracer> contain the emissions of <tracer> from <source> as defined via offemis.nml in units of molecules per square meter per second or molecules per cubic meter per second in case of "2D" or "Nx2D" emissions, respectively; for the different emissions types see [Kerkweg2006]; note that no output can be created for "3D" emissions
  • <source>_<tracer>_vind contain the indices of the vertical layers to which the above emissions were assigned (only present if emissions are actually assigned to multiple layers)


This subsection assumes that ice phase "chemistry" is disabled. In case it is enabled, further output variables will have to be considered, especially in the calculation of total wet deposition fluxes.

  • wetflx_<aggregate> contain the sum of the wet deposition fluxes of the components that make up the <aggregate> in the chemical mechanism of SCAV; these are only available for the <aggregate>s "sulfate", "nitrate", and "ammoni"; for total wet deposition fluxes of the <aggregate> the corresponding wetflx_aer_*_<tracer> (see below; e.g., all wetflx_aer_*_NH4_<modename>, where <modename> is defined as in * have to be added
  • wetflx_sum_<aggregate> contain the time integrals of the wetflx_<aggregate>
  • wetflx_ls_HCl_l, wetflx_ls_Clm_l, wetflx_cv_HCl_l, and wetflx_cv_Clm_l have to be summed to obtain the chloride wet deposition flux that corresponds to the wetflx_<aggregate> (again, if you are interested in the total chloride wet deposition flux you have to add the wetflx_aer_*_Cl_<modename>, see below)
  • wetflx_aer_<cloud>_<tracer>, where <cloud> is either "ls" or "cv" for large-scale or convective clouds, respectively, contain the wet deposition fluxes of the aerosol tracers requested in the CPL namelist in scav.nml that did not end up in the corresponding wetflx_<cloud>_<species>_l, which are only created for species included in the SCAV liquid chemistry mechanism (described above only for chloride; note that the wetflx_<aggregate> described above are sums of the corresponding wetflx_*_<species>_l); hence, if there are any wetflx_*_<species>_l that correspond to a <tracer> base name (see * these have to be added to the sum of the wetflx_aer_*_<basename>_<modename> over all <modename> (defined as in * in order to obtain the total wet deposition flux of the component described by the <basename>; for an example see below
  • frac_evap_<hydro>_<cloud>_<tracer>_mNN contain the fractions of each <tracer> (see * that are assigned to the modes with indeces NN after evaporation of constituents or hydrometeors of clouds of type <cloud>; cloud constituents or hydrometeors are identified by <hydro>, which can take the values "snow" for snow flakes, "rain" for rain drops, "iwc" for in-cloud ice crystals, and "lwc" for in-cloud droplets
To obtain the total SO4 wet deposition flux, sum up wetflx_sulfate, wetflx_aer_ls_SO4_*, and wetflx_aer_cv_SO4_*, which is equivalent to summing wetflx_ls_H2SO4_l, wetflx_ls_HSO4m_l, wetflx_ls_SO4mm_l, wetflx_cv_H2SO4_l, wetflx_cv_HSO4m_l, wetflx_cv_SO4mm_l, wetflx_aer_ls_SO4_*, and wetflx_aer_cv_SO4_*.


  • sediflux_<tracer> contain the sedimentation fluxes of the <tracer>s (see * in units of molecules (or particles, for number tracers) per square meter per second
  • loss_<tracer> contain the time integrals of the above, converted to mass (i.e., [kg m-2]) in case of mass tracers


  • <basename>_<modename> contain the mixing ratios (in [mol mol-1] for mass tracers, in [mol-1] for number tracers) of the aerosol components <basename> in the modes <modename>; aerosol tracer <basename>s are "SO4", "NH4", "NO3", "Na", "Cl", "POM", "BC", "DU", and "H2O" (cf. the Introduction); the <modename>s consist of two letters indicating the size range and particle class, i.e., first letter
    • "k" for Aitken,
    • "a" for accumulation, and
    • "c" for coarse mode,
    and second letter
    • "s" for soluble,
    • "m" for mixed, and
    • "i" for insoluble particles,
    e.g., "SO4_ks" for sulfate in the soluble Aitken mode.