Land Model Experiments
Land Model ExperimentsObjectives
3 Experiments with Land Model
Experiment 1: Basic CLM5 Experiment using Satellite Phenology
Run the land model (CLM5) with satellite phenology (SP) at 1 deg horizontal resolution for 5-days
How can I find out what this compset means?
daily averages (hist_nhtfrq)
make history files contain 5-days + initialization day (hist_mfilt)
What file do I make namelist change in for the land?
What is the command for checking namelists?
Where is it run from?
Reduce the wallclock time.
./xmlchange --subgroup case.run JOB_WALLCLOCK_TIME=0:15:00
When the run is complete, where will the land history data be located?
There will be 6 timesteps in the history file. The first time step is an initialization value. The second is the first daily average.
Experiment 2: Explore Differences between Land Compsets
Run the same configuration as above, but with a different compset.
Where would I look up this difference between this compset and the previous one?
Which files in my two cases should I compare to see how they differ
Finding create_newcase setup
What file tells me the create_newcase command I ran for the first exeriment?
Look at your results
When this run and the previous run are complete, compare the leaf and stem area index (
TSAI), transpiration and canopy and ground evaporation (
FGEV). Note that there will be additional biogeochemistry variables in the second experiment. Use ncview to compare side by side on your screen.
What is the reason for the differences based on what you know about the different compsets?
Experiment 3: Understand Inputs for CLM
Look at the
lnd_in namelist for the first experiment.
Where is the
lnd_innamelist file for the first experiment located?
Find the parameter file specified by the
paramfile namelist item.
What is the name of the file?
This file defines specific information about each of the plant function types in CLM. This is an important input file for the land model.
We will look at this file in Jupyter, specifically the variable
rholvis. This is the visible leaf reflectance for each plant functional type (pft). Its the amount of visible radiation reflected by the different types of plants.
How do I launch Jupyter (or Python in general) from Cheyenne?
This is an input file for the land model. We will change the values in this file and see what that does to the model simulation.
Looking at data in .nc files
What is plant functional type #4?
What is the
rholvisvalue for plant functional type #4?
Creating new input files
Create a new file which changes the visible leaf reflectance of plant functional type #4 (tropical broadleaf evergreen tree) to 0.4.
Write the data to a netcdf file.
Create a new experiment exactly like the first one, but change it to use your new file.
What command creates an exact replica of another experiment?
Where is that command run from?
Changing Land Input files
Where would you change this so your new case will find the new input file?
While the model is building…
What do you think this change did?
What do you expect the the impact to be?
Look at the result
Compare the history output against that generated in the first experiment. Here’s a nice tool for making that comparison:
ncdiff hist_exp2 hist_exp1.nc hist_diff.nc module load ncview ncview hist_diff.nc
How did this change impact your experiment?
Some variables to compare:
Was the impact what you expected?
Why or why not?