Land — ELM

ELM (E3SM Land Model) is E3SM’s land component, based on CLM5. It simulates vegetation, soil, hydrology, and carbon/nitrogen cycles.

Relevant zppy tasks

ts — Time Series Files — generate land time-series files
e3sm_to_cmip — CMIP6 Conversion — convert land output to CMIP6 format
e3sm_diags — E3SM Diagnostics — land diagnostics (e.g., lat_lon set with land variables)
ilamb — Land Benchmarking — land benchmarking against observational datasets
livvkit — Ice Sheet Validation — ice sheet validation (for cryosphere configurations)
global_time_series — Global Time Series Plots — land global time series (via plots_lnd)

ELM time-series configuration

ELM output is typically in monthly elm.h0 files. A typical [ts] configuration for land:

[ts]
active = True

  [[land_monthly]]
  input_files = elm.h0
  input_subdir = archive/lnd/hist
  input_component = elm
  mapping_file = /path/to/map_r05_to_cmip6_180x360_aave.nc
  grid = 180x360_aave
  years = 1:100:10

The input_component = elm setting tells zppy that this is ELM output. This is required for correct processing by e3sm_to_cmip and ilamb.

CMIP conversion for ILAMB

ILAMB requires CMIP6-format data. The e3sm_to_cmip task must be run first. Typical configuration:

[e3sm_to_cmip]
active = True

  [[land_monthly]]
  input_component = elm
  ts_land_subsection = land_monthly

[ilamb]
active = True
land_only = False
e3sm_to_cmip_land_subsection = land_monthly

The ilamb task will depend on both the land and atmosphere e3sm_to_cmip subtasks (unless land_only = True). With infer_section_parameters = True (default), subsection names for land tasks default to land_monthly and atmosphere tasks default to atm_monthly_180x360_aave.

Ice sheet validation with LIVVkit

For cryosphere-enabled simulations, the livvkit task validates the ice sheet component against reanalysis datasets. LIVVkit requires grid-native and regridded climatology files from the [climo] task:

[climo]
  [[land_monthly_climo_native]]
  input_files = elm.h0
  input_subdir = archive/lnd/hist
  input_component = elm
  years = 1:100:20

  [[land_monthly_climo_era5]]
  input_files = elm.h0
  input_subdir = archive/lnd/hist
  input_component = elm
  mapping_file = /path/to/map_r05_to_era5.nc
  grid = era5
  years = 1:100:20

[livvkit]
active = True
climo_subsections = land_monthly_climo_native, land_monthly_climo_era5
icesheets = gis,ais
sets = cmb,smb,energy_era5

Notes for ELM developers

The input_component = elm parameter is used internally by zppy to set the correct processing type (prc_typ) for ncclimo.
ELM variables to process can be set via the vars parameter. Leaving vars = "" processes all variables, which can be slow. Consider specifying only the variables needed for your diagnostics.
The land component mapping file maps from ELM’s finite-volume grid (e.g., r05, ne30pg2) to a regular lat-lon grid.