1. Intermediate model for the annual and global evolution of species
IMAGES:
Intermediate model for the annual and global evolution of species
Description of the model
The reactive carbon cycle
Bottom-up inventories
Model results for CO
NOx chemistry, modelled NOx columns, comparison with SCIAMACHY NO2 columns
Modelling HCHO, comparison with measurements

2. [X]=number concentration of a compound X
Description I
[X]=number concentration of a compound X
« Box model »
N species - N equations
Chemical transport model (CTM)
N x N1 x N2 x N3 equations
N1=number of longitudes
N2=number of latitudes
N3=number of altitudes
operator splitting-less accurate solvers
Extends from the surface to the pressure level of 44 mbar, 40 vertical levels, and 5x5 degree horizontal resolution
Provides the distribution of 48 chemical species (including non-oxygenated organic e.g. C2H6, C2H4, oxygenated e.g. PAN, MPAN, CH3COOOH, carbonyls e.g. HCHO, CHOCHO, CH3CHO, etc., peroxyradicals e.g. CH3O2, etc.)
Short-lived species are not transported in the model

3. Description II
The vertical resolution is higher near the surface
Equation 1 is solved numerically by an operator-splitting technique
Time step = 1 day, except for the three first days of each month : diurnal cycle calculation
Advection is driven by monthly mean climatological fields from ECMWF  short-term wind variability not taken into account – mixing associated with wind variability is comprised in the diffusion term
Turbulent mixing in the PBL  diffusion term
Vertical transport associated with deep convection
Advection  Semi-Lagrangian scheme : suitable for large timesteps

4. Description III
Diffusion equation in 3 dimensions: Kxx, Kyy, Kzz are the zonal, meridional and vertical diffusion coefficients, solve with a implicit Eulerian scheme in each direction
Horizontal diffusion coefficients : proportional to the deviations of wind fields
Vertical diffusion coefficient : depends on the PBL height
Kxx = m2/s at mid-latitudes, much smaller values in the tropics, Kyy = factor of two lower than Kxx
Kzz values are sufficiently high to allow for rapid exchanges of mass between the surface and the free troposphere

5. Description IV Deep convection : Treated as an 1-dimensional process
Assumption : Ascending motions transport air from the boundary layer to the free troposphere, while subsidence transports air from each level to the adjacent lower level only  derive probabilities from the updraft densities
Use updraft fluxes from the ECMWF analyses
Chemistry : P is the photochemical production and beta the loss rate
For transported species, the quasi-steady state approximation is used, or an Eulerian appoximation when beta is close to zero, for short-lived equilibrium is assumed, iterative procedure

6. Description V 48 long-lived and 20 short-lived species
~200 chemical reactions, ~30 photolytic reactions (Muller and Stavrakou, 2005)
Water vapor, pressure and temperature are specified from ECMWF data
Lumping is used to reduce the number of species, e.g. the peroxy radicals formed from ISOP+OH are lumped into one species (MIM is used)
J’s are interpolated from values calculated offline using a radiative model
J’s depend on the sza, ozone column, surface albedo, T, clouds, and z
Heterogeneous reactions on sulphate aerosols N2O5 + SO4  2 HNO3 + SO4, NO3+SO4 HNO3+SO4, HO2+SO40.5 H2O2+SO4, (cloud droplets) N2O5 2HNO3  under construction
Wash-out parameterization : uses large-scale and convenctive precipitation from ECMWF, 3-d cloud cover fields and much more  under construction
The model is parallelized with OpenMP at 95%. It runs on 2,4,8, and 16 cpus. On the 5x5 grid : 20 min for 1-year simulation

7. The reactive carbon cycle (units: Tg C/year)
deposition
deposition 85 30 OH
OH, hv OH
CH2O
CO2 CO
CH4
1100
570
360
CO2
340
deposition
100
OH,O3 80
NMVOC (non-methane volatile organic compounds)
250
SOA
200 50
700
100

8. Bottom-up emissions
Global total : 27 Tg N/yr, source: EDGAR v3.3

9. Global total : 8 Tg N/yr, Yienger and Levy, 1995
Global total : 3 Tg N/yr, Price et al, 1997, Pickering et al, 1998

11. Bottom-up biogenic emissions
MEGAN model coupled with the MOHYCAN canopy model
driven by ECMWF fields and accounts for leaf age, soil moisture stress, and past temperature radiation levels (Muller et al., 2008)
A newly developed inventory for biogenic emissions is used in this study. It is based on the MEGAN model coupled with a canopy
environment model. A detailed description of the inventory is given in the Muller et al., 2007 (not yet submitted) – ask for the preprint if interested ! –
and the files at a resolution of 1 degree can be downloaded at this website. Zonally and monthly averaged biogenic emissions from 1995 to 2006 of
this database are shown here, and comparison with the Guenther et al., 1995 is provided in the last panel. Large interannual variability and lower
emissions by up to 30% with repspect to the G95 database are the two main features of the new inventory.

12. IMAGES results : CO Mixing ratio

13. NOx : role, surces and sinks+
NO2 NO
+ O  O3
RO2 O3 + +
N2O5
H2O
CO, VOC, O3 OH
HO2
2HNO3 O3
HNO3

14. Annually averaged modelled vs. observed NOx column - 2003
SCIAMACHY NO2 column

15. HCHO chemistry, sources and sinks
The most abundant carbonyl in the atmosphere
Short-lived - lifetime on the order of a few hours
Directly emitted from fossil fuel combustion and biomass burning
Also formed as a high-yield secondary product in the CH4, and NMVOC oxidation
CH4
NMVOC
HO2 OH OH
CH3OOH
CH3O2
RO2 OH NO
HCHO OH
CO+2HO2
CO+HO2+H2O
CO+H2
deposition

16. Annually averaged prior NMVOC emissions from biomass burning and biogenic sources - 2003

17. Annually averaged modelled vs. observed HCHO column - 2003
SCIAMACHY HCHO column

18. Impact of NMVOCs on O3 mixing ratios
July 1997
without NMVOCs
with NMVOCs