Skip to content
Alistair Adcroft edited this page Oct 17, 2022 · 4 revisions

Notes only for sessions on the Phillips model

To run the Phillips model longer and get some meaningful output in a short time, add the following lines to ocean_only/Phillips/MOM_override:

#override DT = 3600.
#override DAYMAX = 100.
#override KHTH = 0.
#override RESOLN_SCALED_KH = False
VELOCITY_IC_PERTURB_AMP = 1.0

DT is the baroclinic model time-step. DAYMAX is how many days to run the model. Setting KHTH and RESOLN_SCALED_KH is turning off a parameterization used in Hallberg, 2013. Doing so makes the model closer to a conventional Phillips experiments. VELOCITY_IC_PERTURB_AMP is a parameter unique to the Phillips experiment that we're boosting here to help instabilities to form in a short time.

We ran the model with 4 cores:

mpirun -n 4 ../../build/ocean_only/MOM6

Analyzing MOM6 session

clone the github repo: https://github.com/raphaeldussin/MOM6_tutorial_analysis

and follow the readme to install conda environment.

For the session adapting an experiments

We will step-by-step add the following lines (in addition to the above) to MOM_override:

#override NIGLOBAL = 40
#override NJGLOBAL = 40
REENTRANT_Y = True
#override JET_WIDTH = 1600.
#override JET_HEIGHT = 300.
#override BETA = 0.

#override TOPO_CONFIG = "seamount"
SEAMOUNT_DELTA = 0.7
SEAMOUNT_X_LENGTH_SCALE = 200.
SEAMOUNT_Y_LENGTH_SCALE = 200.

We will also launch a jupyter notebook using the mom6py conda environment created at the bottom of the Tutorials page:

jupyter notebook

URL for FRE session

https://github.com/raphaeldussin/simple_xml