wiki:Documentation/FrenchConfiguration

Version 15 (modified by luyssaert, 9 months ago) (diff)

--

The French configuration(s)

AIM

Set up a configuration for mainland France making use of the best-available boundary conditions and climate forcing for the region. Some intended uses (which are the rational of this task) are described below.

STRATEGY

A configuration called OOL_SEC_STO_FG8nd was created (r6986). Note that configuration 6 and 7 were not yet created but these numbers were reserved for the associated spinup and transient run. OOL_SEC_STO_FG8nd is a copy of OOL_SEC_STO_FG3nd. The plan is to prepare and gradually replace the current boundary conditions and climate forcing until the desired configuration has been reached (see below).

CURRENT CONFIGURATION

  • ORCHIDEE: new driver
  • Vegetation discretization: 15 PFTs with one age class
  • Parameter files: orchidee_pft.def_15pft.1ac
  • Land cover change map: CMIP6/ESA-LUH2v2/historical/15PFT.v2/PFTmap_${year}.nc
  • Climate forcing: 0.5x0.5 degrees annual WFDEI_GPCC forcing between 1979-2009 (about the forcing)
  • Soil map: SOIL/soil_bulk_and_ph.nc, SOIL/soils_param4.nc
  • N-input: NITROGEN/N_FERTILISATION/NMIP/synthetic/historical/Nfer_pasture_${year}.nc, Nfer_cropland_${year}.nc, Nmanure_pasture_${year}.nc, Nmanure_cropland_${year}.nc, CCMI_ndep_nhx_${year}.nc, CCMI_ndep_noy_${year}.nc, NITROGEN/BNF/bnf_1850.nc
  • CO2: CO2/CO2_1860_2012_TRENDY2.txt
  • Management: FOREST_MANAGEMENT/TRUNK_FM/historical/v1/global_forest_management_${year}.nc
  • River basins: ROUTING/routing.nc
  • Restart: Start from scratch

DESIRED CONFIGURATION

  • ORCHIDEE: The files which make up the basic configuration will be prepared for the trunk with the new driver (ORCHIDEE 4.x). These files should be back compatible with ORCHIDEE 3.x and Tags 2.x
  • Vegetation discretization: a single land cover map with the current 15 default PFTs will be prepared for a recent year. The forest, grassland and cropland MTCs can be refined in species-specific PFTs if required by the application.
  • Parameter files: the parameter files will depend on the number of PFTs that was used to discretize the vegetation. For ORCHIDEE 4.x parameter files can be managed through the scripts and database stored in config/ORCHIDEE_OL/MAKE_RUN_DEFS.
  • Land cover change map: Each project will have to prepare its own land cover change maps based on the number of PFTs and the time frame of the simulations, i.e., historical or future.
  • Climate forcing: In the short term the configuration will make use of the SAFRAN forcing. In the future higher resolution forcing produced by ORCHIDEE-WRF could become available.
  • Soil map: INRAE soil map will be used. See discussion
  • Management: each project will have to prepare its own management maps based on the number of PFTs and the objectives of the study.
  • River basins: not working yet at the SAFRAN resolution. For the time being we could store high resolution run-off information through the history files. Run-off could be calculated in post-processing.
  • Restart: each project will have to prepare its own management maps based on the number of PFTs and the objectives of the study.

To do

  • GM: Ask Nicolas Viovy whether his files are for his driver or standard driver.
  • PP: Prepare SAFRAN for 2.1new driver (Philippe will ask Vlad for scripts/insights). Make sure it is compatible with the current and new driver.
  • AD: High resolution basin information needs to be prepared (this is already on Agnes’s to do list). Conversion from lambert-projection to lat/lon.
  • PP: CITEPA ask for their land cover map.

Discussion points

Land cover change map

CITEPA could contribute as they have “made” specific land cover maps for France (annual) based on different products (ESA-CCI, a map from CESBIO, Corinne LC, ..)

Climate forcing

Adjust SAFRAN to the new driver

Soil map

Following email exchange with ADEME, Manuel Martin from INRAE infosol d'Orléans proposes to use the soil map from RMQS-RU for the texture. Need to start on time because this map requires a user agreement. The same product should also contain information on the water holding capacity. These maps have been described in:

Dobarco, M. R., Cousin, I., Le Bas, C., & Martin, M. P. (2019). Pedotransfer functions for predicting available water capacity in French soils, their applicability domain and associated uncertainty. Geoderma, 336, 81-95. https://doi.org/10.1016/j.geoderma.2018.08.022

Dobarco, M. R., Bourennane, H., Arrouays, D., Saby, N. P., Cousin, I., & Martin, M. P. (2019). Uncertainty assessment of GlobalSoilMap? soil available water capacity products: A French case study. Geoderma, 344, 14-30. https://doi.org/10.1016/j.geoderma.2019.02.036

Agnes is also aware of yet another paper about a French soil map but it is not clear whether this is the same map or not

V.L. Mulder, M. Lacoste c, A.C. Richer-de-Forges & D. Arrouays (2016). GlobalSoilMap? France: High-resolution spatial modelling the soils of France up to two meter depth. Science of the total environment, 573, 1352-1369. ORCHIDEE 2.2 (r6950) has two approaches of reading soil information: (a) Read soil texture maps from USDA and derive the required variables as coded in constantes_soil_var.f90 (b) Read a map for each of the seven soil hydraulic parameters required by ORCHIDEE (r6950 still needs to be integrated in the trunk) Whatever option is used a soil texture map is needed. When using the INREA Infosol map we should know whether the map described the soil texture at the surface or at a depth of 30 cm

The key soil parameters:

  • clay content
  • ks (hydraulic conductivity),
  • alpha and n (de van Genuchten parameters)
  • mcs (saturation),
  • mcr (residual),
  • mcw (wilting point),
  • mcf (field capacity)

Note that the latter two can be calculated from the former 5 parameters. The water holding capacity is calculated as the product of (mcf-mcw), mcs, and the soil depth.

Suggestion [AD]: map the dominant soil texture class at the SAFRAN resolution. Also map the derived soil parameters (constantes_soil_var.f90) and compare with maps showing these parameters.

Intended use(s)

ADEME

  • Involved: Nicolas Vuichard, Jina Jeong, Guillaume Marie and Sebastiaan Luyssaert
  • Short description: The ADEME project studies the impact of natural disturbances and forest management on future forest growth in France.

Configuration spinup

  • ORCHIDEE 4.x: sechiba, stomate and sapiens
  • Vegetation discretization: tree species level, three diameter classes and three or four age classes
  • Restart: none
  • Land cover change map: to be created based on the NFI age distribution map, NFI species distribution map, and the LUHv2 or another LCC product.
  • Climate forcing: medium resolution (~ 8x8 km) no other specific requirements. Currently looking into SAFRAN data 1986 to 2018.
  • Soil map: if available the soil map from a previous ADEME project. If not any soil map could do.
  • Forest management map: for a recent year (default or based on NFI)

Configuration transient

  • Included in the spinup thanks to a dedicated land cover change map. This method only works when we try to match the forest age in a recent year. A present day forest age map could be created based on the NFI data. With this information we know when the forest was planted. We can then look at the LUH maps to see whether the pixel was forest or not. If it was a forest we keep the same forest if it was a grassland or cropland we change it. This gives the land cover map than will be used for the spinup. Following the spinup a transient run will account for the LCC in the correct year. That method will result in the correct age distribution of the forest in France in a recent year.

Configuration simulation experiments

  • ORCHIDEE 4.x: sechiba, stomate and sapiens
  • Restart: from spinup
  • Land cover change map: none
  • PFT map: based in NFI data (last map of the Land cover change maps that need to be prepared for the spinup)
  • Climate forcing: downscaled AR5. medium resolution (~ 8x8 km). We will have an inconsistency between the spinup and the future simulations no matter which forcing we use.
  • Soil map: if available the soil map from a previous ADEME project. If not any soil map could do.
  • Species change maps: based on the scenarios France (this is part of the simulation experiment)
  • Management change maps: based on the scenarios France (this is part of the simulation experiment)

Services climatiques IPSL, Explore2 (MTES), BLUEGEM if funded by Belmont Forum

  • Involved: Agnès Ducharne, Jan Polcher, Frédérique Chéruy, Philippe Peylin, Bertrand Guenet, Philippe Ciais
  • Short description: All three projects have a focus on water resources and their response to climate change and anthropogenic pressures, especially from the agricultural sector.

Configuration simulation experiments

  • ORCHIDEE 2.x (3.x foreseen in a second step with Nicolas Vuichard): sechiba, stomate with HR routing, GWF, and irrigation
  • Restart: from spinup
  • PFT map: first step with CMIP6 maps, and attempts to use French agricultural census
  • Land cover change: yes, using standard CMIP6 maps
  • Climate forcing: SAFRAN (~ 8x8 km) + downscaled bias-corrected climate projections from other partners of the projects ; CORDEX output
  • Soil map: first step with Reynolds, but tests with the INRA soil map
  • Management change maps: for agriculture and irrigation, using standard CMIP6 maps (LUHv2), and tailored irrigation
  • Model calibration : with ORCHIDAS over the historical period, to match hydrological observations

Produce forcing to run ORCHIDEE at km-scales

  • Involved: Jan Polcher, Nicolas Vuichard, Nathalie De Noblet, …
  • Short description: To be able to use ORCHIDEE over France at an appropriate resolution we also need the adequate forcing. Not only in terms of spatial resolution but also on the quality of the diurnal cycle of rainfall. I.e. the intensity of rainfall needs to be correct so that all the issues with ORCHIDEE’s water cycle come to light and we can explore our ability to simulate extreme events. This would mean running RegIPSL (WRF+ORCHIDEE) forced with ERA5. It would provide a dynamically downscaled version of the re-analysis which can serve for model development ... and in the end see how these improvements feedback to the atmosphere.