Ocean — MPAS-Ocean

MPAS-Ocean is E3SM’s ocean model, based on the Model for Prediction Across Scales (MPAS) framework. It uses an unstructured mesh that allows variable horizontal resolution.

Relevant zppy tasks

mpas_analysis — Ocean and Sea-Ice Analysis — ocean and sea-ice analysis (primary task)
global_time_series — Global Time Series Plots — ocean global time series plots

MPAS-Analysis for ocean

MPAS-Analysis is the primary tool for MPAS-Ocean diagnostics. It operates on MPAS-Ocean’s native unstructured mesh. The [mpas_analysis] task in zppy calls MPAS-Analysis automatically.

Typical [mpas_analysis] configuration:

[mpas_analysis]
active = True
walltime = 06:00:00
shortTermArchive = True
ts_years = 1:100:50
climo_years = 1:100:50
enso_years = 1:100:50
generate = all, no_landIceCavities, no_BGC, no_icebergs, no_min, no_max

Key MPAS-Ocean-specific parameters:

mpaso_nml: name of the MPAS-Ocean namelist file (default: mpaso_in)
stream_ocn: name of the MPAS-Ocean streams file (default: streams.ocean)
PostMOC: whether to post-process MOC (meridional overturning circulation) data

Model-vs-model ocean comparisons

MPAS-Analysis supports comparing two model simulations. In zppy, this is configured via reference_data_path:

[mpas_analysis]

  [[main_run]]
  ts_years = 1:100:50
  climo_years = 1:100:50

  [[mvm_run]]
  ts_years = 1:100:50
  climo_years = 1:100:50
  reference_data_path = [[main_run]]
  # or: reference_data_path = /path/to/prior/zppy/output

See the mpas_analysis — Ocean and Sea-Ice Analysis page and the tutorial for a complete model-vs-model example.

Global time series ocean plots

The [global_time_series] task can produce ocean time series plots such as:

change_ohc — ocean heat content change
max_moc — maximum meridional overturning circulation
change_sea_level — global mean sea level change

These require access to MPAS-Ocean output. The input_subdir in [global_time_series] should be set to archive/ocn/hist (the default).

Notes for MPAS-Ocean developers

The generate parameter in [mpas_analysis] controls which MPAS-Analysis tasks to run. Use no_* filters to exclude irrelevant analyses for your simulation (e.g., no_BGC if biogeochemistry is not enabled, no_landIceCavities if land ice cavities are not used).
The environment_commands for [mpas_analysis] must be consistent across all year-range runs. Mixing environments can cause MPAS-Analysis to fail.
MPAS-Analysis caches intermediate results (cache = True by default). Set purge = True to clear previous output and start fresh.