ECHAM model setup

  1. grid interpolation

    2. Use CDO remap functions: remapcon, remapbil, remapnn ..... (see CDO manual for details) or NCL regriding functions.

    example: interpolate other resolution to t159 resolution use CDO

    remapbil cdo remapbil, t159grid ifiles ofiles

  2. modify a sub region
    • create a mask file, where area to be changed with mask =1, else mask=0
    • use NCL function "where" or "mask", or CDO function "ifthenelse" to replace variables just for the target region.
  3. reconstruct paleo variables (SST, SIC, OROMEA ...) using anomaly method

    4. Anomaly method is a method to use scientifically other model outputs as a target model boundary condition to avoid adding in the specific model system error.

    e.g. use atmosphere-ocean coupled model T31 paleo SST and SIC outputs as ECHAM T159 ocean boundary condition.

    • create file1: a T31 anomaly file ( T31 paleo - T31 PI) use e.g. CDO function "sub"
    • interpolate file1 to T159 grid file2 use e.g. CDO function "remap**"
    • create file3: paleo T159 by adding file2 to PI T159 use e.g. CDO function "add"
  4. manual adjustment of paleo topography

    5. e.g. close hudson bay

    • create a hudson bay mask where topography need to be modified
    • inside this mask, give the OROMEA to e.g. 5 m where ocean was.
  5. create ECHAM topography related variables: parameterization of ECHAM topography

    6. Once the paleo T159 topography file is created, proceed the following steps to create topography related ECHAM input variables: ORO* and GEOSP

    • interpolate T159 topography file to a higher resolution file4: e.g. to a 0.15 resolution file use CDO function "remap*"
    • run ECHAM parameterization script "echo_topo_parameterize_t159" to create ORO* and GEOSP variables using file4. NOTE: this scripts is modified from a source script originally created by Dr. Ran Feng from university of Michigan.

    echam_topo_parameterize_t159

    • copy those variables to ECHAM input file in a global or regional scale
  6. paleo vegetation reconstruction
    1. Easy approach : from Jasbach outputs or AWI T31 input files, use CDO remap functions.
    2. Difficult approach: use paleo biome map from PMIP MH and LGM biome map, gaps filled with model output Arnold et al. 2009. (this method is not recommended, GCM coupled with dynamic vegetation model will be a better approach)

    7. PMIP MH and LGM biome source data:

    LGM_pmip3.csv

    MH_pmip3.csv

    Arnold et al. 2009, MH and LGM model biome outputs:

    bme_6k.txt

    bme_21k.txt

    Data processes:

    1. summarize ECHAM5 modern biome type based on ECHAM5 inputs and table 2 in Report. 336 and 289. Algorithm: for certain grid box, search through biome types in table 2, and assign the one biome type to the grid box with an least error (sum of the absolute difference of vegetation related variables in ECHAM5 input files to the value gave in table 2 ). repeat this step for all grid boxes.

      2. csv file for table2:

      table2.csv

      code in NCL:

      ncl_make_echam_surface_calculation

    2. convert PMIP3 data and Arnold model output to ECHAM5 grid. Use NCL function triple2grid, or CDO remap functions.
    3. translate PMIP and Arnold MH and LGM biome type to ECHAM5 biome type (by Jian Ni from AWI)

      4. translation_pmip2_to_echam5

    4. create paleo biome map in ECHAM5 grid. Use PMIP observational data when available, missing data will be replaced with model outputs of Arnold et al. 2009.
    5. compare paleo biome type with modern of a target grid box, if different, search through modern biome type and find a the nearest grid box in modern which has the paleo biome type, and then take all vegetation related variable of that grid box and assign or modify them to the target grid box. Repeat this steps for all the grid boxes.

      6. ncl_modify_veg_input_2014

  7. final control
    • max min of all variables: if they exceed the default range (the default max min are assumed to be the values which ECHAM can handle), replace with values > default max with default max, replace values < default min with default min
    • missing values: if there are missing values, replace them with e.g. closest neighbor
    • ocean: where SLM = 0, all variables should describe the condition for ocean (if there are mismatch, replace with ocean data)
    • glacial: where GLAC=1, all variables should describe the condition for glacial (if there are mismatch, replace with glacial data)
    • others, visual check, manual adjustment.

NOTE: Contact Jingmin Li (This email address is being protected from spambots. You need JavaScript enabled to view it.) if things are not clear.

References:

  • Arnold, L., F.M. Breon and S. Brewer, 2009, The earth as an extrasolar planet: the vegetation spectral signature today and during the last Quaternary climatic extrema, International Journal of Astrobiology (8), 81-94. doi: 10.1017/S1473550409004406.
  • Hagemann, S. 2002, An improved land surface parameter dataset for global and regional climate models. Report No. 336, Max Planck Institute for Meteorology.
  • Hagemann, S., M. Botzet, L. Duemenil and B. Machenhauer, 1999. Derivation of global GCM boundary conditions from 1KM land use Satellite data. Report No. 289, Max Planck Institute for Meteorology.