Overview

Teaching: min
Exercises: min

Questions

• Land Model Experiments

Objectives

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

compset: I2000Clm50Sp

resolution: f09_g17_gl4

Compset information

How can I find out what this compset means?

https://www.cesm.ucar.edu/models/cesm2/config/2.1.1/compsets.html

Namelist changes:

• daily averages (hist_nhtfrq)

• make history files contain 5-days + initialization day (hist_mfilt)

Namelists

What file do I make namelist change in for the land?

What is the command for checking namelists?

Where is it run from?

Configuration changes:

Reduce the wallclock time.

./xmlchange --subgroup case.run JOB_WALLCLOCK_TIME=0:15:00

History Files

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.

compset: IHistClm50BgcCrop

Compsets

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 (TLAI,TSAI), transpiration and canopy and ground evaporation (FCTR,FCEV,FGEV). Note that there will be additional biogeochemistry variables in the second experiment. Use ncview to compare side by side on your screen.

Interpreting output

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.

Resolved Namelists

Where is the lnd_in namelist 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 rholvis value 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.

Cloning

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: FSRVD,FSRVI,FSR,FSA,FSH,FCSTR,TV,TSA).

Was the impact what you expected?

Why or why not?

Key Points