Running CNP2

[ma595]

This has been added to the merged PR comments, but it warrants being highlighted here too:

These are the changes needed to varlist.json to enable visualisation to work.

 "resp":
  {
    "format": "regular",
    "sourcefile":"/rds/project/rds-5mCMIDBOkPU/rse/spin/data_CNP2/CNP13.global.halfdeg.trendyv10.spinup.1700.MLA.regular_49301231_stomate_rest.nc",
    "targetfile":"/rds/project/rds-5mCMIDBOkPU/rse/spin/data_CNP2/CNP13.global.halfdeg.trendyv10.spinup.1700_13001231_stomate_rest.nc",
    "missing_value":1e20,
    "pool_name_som":["Active","ChemProtect","PhysProtect"],
    "pool_name_biomass":["Leaf","SapAB","SapBE","HeartAB","HeartBE","Root","Fruit","Carbres","Labile"],
    "pool_name_litter":["StructuralAB","StructuralBE","WoodyAB","WoodyBE","MetaAB","MetaBE"],
    "pool_name_microbe":["RmicrAB","RmicrBE","KmicrAB","KmicrBE"],
+   "pool_name_lignin":["StructuralAB","StructuralBE","WoodyAB","WoodyBE"],
    "sub_item":["Cpool","Npool","Ppool"],
    "additional_vars":["npp_daily","npp_week","gpp_week","maxgppweek_lastyear", "gpp_daily"],
    "cnp":3,
    "npfts":14,
    "dim":
    {
      "som":["pft","pool"],
      "biomass":["pool","pft"],
      "litter":["pft","pool"],
      "microbe":["pft","pool"],
+     "npp":["pft"],
+     "lai":["pft"],
+.    "lignin":["pft"]
    },
    "sect_n":
    {
      "som":1,
      "biomass":1,
      "litter":2,
      "microbe":2,
+     "npp":1,
+     "lai":1,
+     "lignin":2
    },

[dsgoll123]

Currently I am testing if we can work without lignin. If lignin state variables need to predicted we need to add them to all configurations.

I am not sure why lai and npp is here.

[dsgoll123]

The test with CNP v1.3 lead similar large biases in re-run as in Sun et al GCB. Is it because lignin is not predicted ? Might be worth a test