ORCHIDEE - BRIC : Biogenic ReactIve Compounds in ORCHIDEE

• Biogenic emissions of Volatile Organic Compounds (VOCs) from vegetation and Nitrogen Oxides (NOx) from soils are now available in the ORCHIDEE trunk version (SVN HEAD)

1. Context

The terrestrial biosphere is the main source of volatile organic compounds (VOCs) into the atmosphere, contributing to more than 80% of global annual emissions of VOCs. About 1000 TgC/yr are therefore emitted by vegetation, mainly from leaves, into the low regions of the atmosphere. In comparison, anthropogenic emissions of VOCs, mostly related to fossil fuel and biomass burning, do not exceed 100 TgC/yr. Isoprene (alkene: C5H8, global emissions: 400-600 TgC/yr) and monoterpenes (C10H16, global emissions ~130 TgC/yr) emissions dominate biogenic VOC emissions but several campaigns brought to light significant biogenic emissions of methanol, acetone, acetaldehyde as well as formic and acetic acids. Nitrification and denitrification processes in soils are also a significant source of nitrogen oxides (NOx). Biogenic compounds such as BVOCs and NOx are usually very reactive molecules, with lifetime ranging from a few minutes to hours for most of them in the troposphere, and are several crucial for atmospheric chemistry processes (ozone cycle, secondary organic aerosol formation, etc.). The vegetation type (boreal forests, tropical forests, grasses…), climatic conditions (temperature, radiation, humidity) as well as human activities (land-use change, deforestation, reforestation) strongly determine the biogenic emission level of those compounds. Having parameterizations directly implemented in a vegetation model to calculate biogenic emissions of VOCs and NOx is therefore of strong interest, giving the possibility to quantify and study the evolution of those emissions in relation with changes in climate, vegetation distribution and ecosystem growth.

2. Biogenic emissions module development (PhD Juliette Lathière, LSCE, 2005)

For more information, especially for a detailed description of parameterizations, see the following references:

J. Lathière (2005), Evolution des émissions de composés organiques et azotés par la biosphère continentale dans le modèle LMDz-INCA-ORCHIDEE, Thèse de doctorat, Université Paris VI.

10. Lathière, D.A. Hauglustaine, N. De Noblet-Ducoudré, G. Krinner et G.A. Folberth (2005), Past and future changes in biogenic volatile organic compound emissions simulated with a global dynamic vegetation model, Geophysical Research Letters, 32, doi: 10.1029/2005GL024164.

10. Lathière, D.A. Hauglustaine, A. Friend, N. De Noblet-Ducoudré, N. Viovy, et G. Folberth (2006), Impact of climate variability and land use changes on global biogenic volatile organic compound emissions, Atmospheric Chemistry and Physics, 6, 2129-2146.

2.1: General code description and location in ORCHIDEE

The core of the code for biogenic VOC and NOx emissions is located in ‘diffuco_inca’, a new subroutine created for this purpose in ORCHIDEE.

TO BE FURTHER COMPLETED

2.2: Related flags

Several flags have been created to activate the different options available for biogenic emissions. By default, all those flags are set to FALSE.

• DIFFUCO_OK_INCA: if ‘TRUE’, allow calculation of VOCs and NOx biogenic emissions, using the subroutine diffuco_inca.

• LEAFAGE_OK_INCA: if ‘TRUE’, the impact of leaf age on isoprene and methanol emissions is considered, otherwise, related impact functions set to 1.

• CANOPY_EXTINCTION: if ‘TRUE’, the extinction of radiation inside the canopy is considered by splitting the leaf area between sunlit and shaded leaves. Radiation received by sunlit leaves on one hand, and sunlit leaves on the other hand, is then estimated and used to calculate isoprene emissions (isoprene being the only BVOC considered to vary with radiation). Otherwise, all leaves are considered to receive the same radiation, and no partition between sunlit and shaded leaves is done.

• CANOPY_MULTILAYER: if ‘TRUE’, a vertical distribution of leaves and extinction of radiation through the canopy is taken into account to calculate isoprene emissions

• NOx_RAIN_PULSE: if ‘TRUE’, the increase in NOx emissions from soils, occurring when a dry surface receives a sufficient amount of precipitation, is considered (or so-called ‘pulse’ of emissions). Otherwise, the related pulse function is set to 1.

• NOx_BBG_FERTIL: if ‘TRUE’, the impact of biomass burning on NOx soil emission increase is taken into account, based on the ‘…’ map, read and interpolated at the beginning of the run.

• NOx_FERTILIZERS_USE: if ‘TRUE’, the impact of fertilizers use on NOx soil emission increase is considered, and the related map ‘…’ is read and interpolated at the beginning of the run. Otherwise, the related function ‘…’ is set to 1 and the map…

3. Merge in ORCHIDEE (Nicolas Vuichard and Didier Solyga, 2011-2012, LSCE)

3.1: Adaptation to the ORCHIDEE AR5 version (Nicolas Vuichard, 2011)

The inclusion and required adaptation of the biogenic emissions code to the ORCHIDEE AR5 version has been carried out by Nicolas Vuichard.

TO BE FURTHER COMPLETED

3.2: Merge in the trunk version (Didier Solyga, 2012)

The merge in the trunk version was carried out by Didier Solyga, right after the ORCHIDEE 1.9.6 tag (revisions 890-906). The modules modified by the merge are:

• ORCHIDEE : intersurf, diffuco, sechiba, slowproc, thermosoil, stomate, pft_parameters, constantes
• ORCHIDEE_OL : dim2driver, readdim2, teststomate

Several problems in the code were fixed:

• Memory allocation: The variable 'pulse' was always used but allocated only if the flag control%ok_pulse_NOx was set to 'TRUE'.

• LAI: This problem is not related to biogenic emissions especially but activating this part of the code gave the opportunity to underline this problem. Indeed, when no restart file is used, the LAI is first calculated in slowproc_lai (even when no map is considered) and this value is then used in the subroutine diffuco_inca. The LAI is set to 0. just before the call to 'histwrite' in stomate. In the meantime, the isoprene fluxes are calculated with a non-zero LAI value, which would lead to non-zero isoprene emission flux, even when no vegetation exists. The LAI is now set to 0. when there is no restart file.

• LMDz interface: The variable 'mean_sinang', required to calculate isoprene emissions, is calculated in diffuco_inca but is also a variable provided by LMDz. To fix temporarily this problem, the variable 'mean_sinang' is set as an optionnal argument in the LMDz interface (intersurf_gathered), and put as the last argument in intersurf_main, set to 0. when LMDz is used in coupled mode. On a longer-term, a special care will have to be given when developping and running ORCHIDEE in coupled mode with LMDz, especially for the variables related to radiation and needed to calculate biogenic emissions: part of all of the radiation scheme implemented in diffuco_inca could be modified, using variables provided by LMDz.

• Activating biogenic emissions required to read the time step of the forcing file to calculated the variable 'nbre_precip', which was originally performed in diffuco_inca, without checking if being in coupled mode. To fix this, the flag OFF_LINE_MODE was moved from the module 'intersurf' to the module 'constantes' and reading the time step is then protected by this flag. In coupled model, the variable 'nbre_precip' is set to 1. This value will have to be checked and possibly changed to another value.

Using teststomate:

• The variable frac_age_x is added to adapt the biogenic emissions ORCHIDEE version to teststomate. This variable is not actually used in teststomate but is rather a formal implementation required to run the programme.

Other changes performed:

• Messages implemented for tables memory allocation.

• The variable 'veget_max10a13' was replaced by the generic variable 'veget_max_nowoody' to take into account the externalisation.

• Cleaning with some parameters being removed (splwORCH, d3_pulse,w1_pulse, w2_pulse, quiet...)

• The documentation for biogenic VOC and NOx emissions included.

• Parameters originally located in 'constantes_inca.f90' are now in 'pft_parameters.f90' (for emission factors, depending on vegetation type) and 'constantes.f90' (for emission activity factor depending on leaf age classes).

• Parameters for equation coefficients were added in the code to supply values written directly in the equations (CT1, CT2, TS, TM, alpha, CL1, beta, lai_threshold).

Several tests were then performed to validate this version:

• Test offline on Obelix : OOL_SEC_STO configuration, 1 proc, 1982, ncc forcing, 1M, daily outputs, no activation of biogenics

• Test offline on Obelix : OOL_SEC_STO configuration, 1 proc, 1982, ncc forcing, 1M, daily outputs, with and without reading the forcing required for biogenics.

• Test online with LMDZ coupled on Titane : LMDZ_ORC, 32 procs, 1980, 1Y, monthly outputs, biogenic not activated (testing the interface)

• Test online with LMDZ coupled on Titane : LMDZ_ORC, 32 procs, 1980, 1Y, monthly, without reading the forcing required for biogenics.

4. Perspective

• So far the different flags for biogenic emissions are described in orchidee.default, and should be also included in the ‘run.card’ file so that any user can set them depending on the run purposes.

• The code related to biogenic VOC and NOx emissions is currently mainly located in diffuco.f90. It could be suitable to gather all developments performed in ORCHIDEE, and dealing with biosphere-atmospheric chemistry interactions, in one new module. To be discussed.

• This work is the first step dealing with the interactions between the terrestrial biosphere and the atmospheric chemical composition. The next step, in progress, will be the coupling between ORCHIDEE and the chemistry-transport model LMDzINCA: see the page.