-> Aerosol Comparisons between Observations and Models



Protocol Further Details on AeroCom protocol

1) Enough info on your model grid (area and vertical coordinate info)
to reconstruct budgets, zonal plots and area means
Communicate with us on the vertical grid info available for your model.

2) Info on any specific deviation from the protocol and/or units given there.
Establish a corresponding ascii list.

3) Info on exact units for Sulfur and Carbon compound
We prefer kgS and kgC instead of kg !!!
Please indicate also the Carbon-to-ParticulateOrganicMatter conversion factor.

4) Additional info on Sulfur/Sulfate emissions
Did you include gas-phase and liquid-phase sulfur production?
What do you mean by "emission of SO4"?
See also new suggestions in protocol-addup for differentiating emissions

5) Info on eventual Carbon or Sulfur production terms above surface level.

6) A check on whether you did use the unit of kg/m2 (or eventually kg/kg or kg/m3) for
Monthly vertical resolved dry mass loads etc.
Did you use kg/m2/month? (...we will take care of correcting for months of different length!)

7) Aerosol optical depth per species for aerosol internal mixtures
A check on how you compute the contribution of a given aerosol species to total aerosol optical depth if you have internal mixtures in the model:

Procedure recommended during AEROCOM workshop: Compute volume fraction of aerosol species in aerosol particle volume (without water!!) and retrieve with that fraction the aerosol optical depth for a given species. Apply the following densities for the major species (Dust = 2650 kg/m3 / Sea salt = 1600 / Sulfate = 1769 / Black Carbon = 1500 / Particulate organic matter = 1500 )

( we recognise that this might not be possible within your time frame)

1) A list of your variable names (which correspond to protocol variables),
(at best put into the excel file of the protocol, see above)

2) That you apply the protocol_addup to experiment B
and also to experiment A (if possible)

3) Further info on your model
Is there a model website to connect to?
Info on aerosol dynamics, transport and removal process formulation.
(Stefan Kinne will collect the info and produce overview tables)

For info: How processing of model data files is done

NCO data pre-processing
Your variables are renamed by an nco script to fit our nomenclature (see protocol),
(NCO web site !!)

NCO commands are also used to produce eg annual or monthly averages.

The indexing of the model grid is different for different model,
this is taken care of by these scripts.

IDL pre-processing:
A setup routine is prepared for each model.
Any variable not having the units asked for in the protocol is converted.
Please find details on the calculations below.
Sometimes arrays need to be transposed to fit idl-array conventions.

IDL Calculations

From the 3D Field of Load [kg/m2]: Sum over the height

Some of you have given a 3D Conc field instead of a 3D Load field.
From the 3D Field of Conc [kg/m3]: Integration over the height

conversion from 3D Field of Load [kg/m2] to 3D field of Conc [kg/m3]:
multiplication with the grid box height dh: Conc=Load*dh

conversion from Conc to mass mixing ratio MMR: division with density of air rho

Load [kg/m2]: area weighted mean

Conc [kg/m3]: volume weighted mean

Mass Mixing Ratio MMR [kg/kg]: mass weighted mean

Volume Mixing Ratio VMR [m3/m3]: volume weighted mean

Calculation of total load load_tot and fluxes flux_tot in the model domain

height of a grid box dh [m] to calculate the volume from area and height
dh can be calculated from dh=pdel/rho/g
where pdel is a 3D field of the pressure gradient of the grid box,
rho a 3D field of the density [kg/m3],
g the gravity constant 9.81

rho can be derived from rho=pmid/temp/287
where pmid is a 3D field of the pressure of the grid box [Pa],
temp a 3D field of the temperature [K]

mass per grid box
airmass can be calculated from airmass=pdel/g*area
where pdel is a 3D field of the pressure gradient of the grid box,
area is a 2D field of area of grid box [m2]
g the gravity constant 9.81

horizontal cross sections at certain pressure levels or vertical cross sections of 3D fields
For these plots we need vertical information on grid.

pmid pressure the grid cell center [Pa]
for sigma-p-coordinates pmid=sigma*(p_surf-p_top)+p_top
for hybrid sigma-p-coordinates phalflevel= Ak + Bk*p_surf

height h [m] (optional)
h can be derived from h(k)=h(k-1)+dh where k is the model level index
the height h0 of the lowest level (orogrophy) needs to be known

For info: Review of equivalent formulations to obtain grid info

2D field of area of grid box [m2]

height of a grid box (to calculate the volume from area and height)
3D field dh height of a grid boxes [m]

or 3D field of the pressure gradient of the grid box pdel [Pa],
3D field of the density rho [kg/m3],

or 3D field of the pressure gradient of the grid box pdel [Pa],
3D field of the pressure in the grid box pmid [Pa],
a 3D field of the temperature [K]

3D field airmass of grid box [kg]

or 3D field of the pressure gradient of the grid box pdel [Pa],
3D field of the temperature temp [K]

3D field of pressure [Pa]

for sigma coordinates
or 3D field of sigma
2D field of surface pressure [Pa]
2D field of top pressure [Pa]

for sigma-pressure-hybrid coordinates ???

height (optional)
3D field of height [m]

or 3D field of height of surface level h0 [m]

is an international
science initiative
on aerosols and climate

supported by
EU Framework programmes

Norw. Met Office
Max-Planck Ges.
French CNES