Task 1

Task 1.0 Archive data to NIRD

Transfer your data from Fram/Betzy to NIRD

# copy files from Fram/Betzy to NIRD
# alternative 1, use scp
scp -r /cluster/work/users/<username>/archive/<casename>/
 <username>@login.nird.sigma2.no:/projects/NS2345K/workshop2021/<username>/<casename>/
# For example:
scp -r /cluster/work/users/yanchun/archive/NHISTfrc2_workshop2021/  yanchun@login.nird.sigma2.no:/projects/NS2345K/workshop2021/yanchun//NHISTfrc2_workshop2021
# alternative 2, use rsync
rsync -aP --chown=OWNER:GROUP <path/to/noresm/output> <username>@login.nird.sigma2.no:<path/to/project/storage>
# For example,
rsync -aP --chown=yanchun:NS2345K /cluster/work/users/$USER/archive/NHISTfrc2_workshop2021/ yanchun@login.nird.sigma2.no:/projects/NS2345K/workshop2021/$USER/NHISTfrc2_workshop2021/

See more information on archiving NorESM output at the NorESM documentation

Task1.1 Set up

Option 1: use pre-installed tool

Use pre-installed tool under NS2345K project of NIRD (Recommended)

# Fist, Logon NIRD
$ ssh -l <your_username> login.nird.sigma2.no

# Next, add alias for diag_run (by pasting the following 5 lines on Shell)
cat <<EOF >> ~/.bashrc
if [ -f /projects/NS2345K/diagnostics/noresm/bin/diag_run ];then
    alias diag_run='/projects/NS2345K/diagnostics/noresm/bin/diag_run'
fi
EOF

# source to take effect
$ source ~/.bashrc

# check if diag_run is in your searching path
$ which diag_run

Option 2: install your own copy

If you have no access to NS2345K on NIRD. You can install your own copy.

$ cd ~/
$ git clone https://github.com/NordicESMhub/noresmdiagnostics
$ cd noresmdiagnostics/bin
$ ./linkdata.sh

It will link the data under either:

/projects/NS2345K/www/diagnostics/inputdata/

to your installed copy.

Note, if you need to download the code to your home folder, you should donwload the observational data by bin/dloaddata.sh to a direcory that has large quota, e.g., to /cluster/work/users/$USER, not directly under your home directory. Then link the downloaded data with bin/linkdata.sh

Next, get familiar with the diag_run optins

# run this wraper script without parameters
$ diag_run

Will shows basic usage:

-------------------------------------------------
Program:
/projects/NS2345K/diagnostics/noresm/bin/diag_run
Version: 2.1
-------------------------------------------------
Short description:
A wrapper script for NorESM diagnostic packages.

Basic usage:
diag_run -m [model] -c [test case name] -s [test case start yr] -e [test case end yr] # Run model-obs diagnostics
diag_run -m [model] -c [test case name] -s [test case start yr] -e [test case end yr] -c2 [cntl case name] -s2 [cntl case start yr] -e2 [cntl case end yr] # Run model1-model2 diagnostics
nohup /projects/NS2345K/diagnostics/noresm/bin/diag_run -m [model] -c [test case name] -s [test case start yr] -e [test case end yr] &> out & # Run model-obs diagnostics in the background with nohup
...
...

Task 1.2 Model-obs comparison of a fully coupled simulation

## Compare model to observation
# syntax:
$ diag_run  -m MODEL -c CASENAEME -s START_YEAR -e  END_YEAR -i INPUT -o OUTPUT -w WEBPAGE​
# examples:
# 1, compare NorESM2-LM historical run (years 1985 - 2014) with observations 
$ diag_run -m blom -c NHIST_f19_tn14_20190710 -s 1985 -e 2014 -i /projects/NS9560K/noresm/cases -o /projects/NS2345K/diagnostics/noresm/out/$USER -w /projects/NS2345K/www/diagnostics/noresm/$USER
# 2, compare NorESM2-LM piControl run (years 1735 - 1764, equivalent to 1985 - 2014), only diagnose the ocean component, and omit the -o and -w options (default to the above settings).
$ diag_run -m blom -c N1850_f19_tn14_20190621 -s 1735 -e 1764 -i /projects/NS9560K/noresm/cases

Challenge

• Use your own finished experiment, and chose the component you want to diagnose
• You should specify the -i, -o and -w option if they are not default paths. Such as, on NIRD:

  -i /projects/NS2345K/workshop2021
  -o /projects/NS2345K/diagnostics/noresm/out/$USER
  -w /projects/NS2345K/www/diagnostics/noresm/$USER
  # or
  -w /projects/NSxxxxK/www
  

  Make sure these two directories exist. An example:

  $ diag_run -m blom -c N1850frc2_workshop2021 -s 1 -e 10 -i /projects/NS2345K/workshop2021/yanchun
  

For those dont’ have succefully model run, ideally longer than 2-years.

You can find sample cases, under: /projects/NS2345K/workshop2021/yanchun

• N1850frc2_workshop2021 (model years, 1-10)
• NHISTfrc2_workshop2021 (model years, 1850-1859)

For those have access to NS2345K and NS9560K on NIRD, there are plent of CMIP6 experiments:

• /projects/NS2345K/noresm/cases
• /projects/NS9560K/noresm/cases

Find out where is the casename and the location of each experiment: https://noresmhub.github.io/noresm-exp/intro.html

Finally, after the diag_run is finished running, browser the result at http://ns2345k.web.sigma2.no/diagnostics/noresm/<usernname>

Task1.3 Model-model comparison

Compare model to model

# Syntax:
$ diag_run  -m cam -c1 CASENAEME1 -s1 START_YEAR1 -e1  END_YEAR1 -c2 CASENAME2 -s2 START_YEAR2 -e2 –END_YEAR2 -i1 INPUT1 -i2 INPUT2 -o OUTPUT –w WEBPAGE
# example, compare NorESM2-LM historical run, years, 1985 to 2014 to piControl (years, 1735 - 1764)
$ diag_run  -m cam -c1 NHIST_f19_tn14_20190710 –s1 1985 -e1 2014 –i1 /projects/NS9560K/noresm/cases -c2 N1850_f19_tn14_20190621 -s2 1735 -e2 1764 -i2 /projects/NS9560K/noresm/cases

Challenge

User your own experiment!
For example:

$ diag_run -m blom -c1 NHISTfrc2_workshop2021 -s1 1850 -e1 1859 -i1 /projects/NS2345K/workshop2021/yanchun -c2 N1850frc2_workshop2021 -s2 1 -e2 10 -i2 /projects/NS2345K/workshop2021/yanchun -o /projects/NS2345K/diagnostics/noresm/out/$USER -w /projects/NS2345K/www/diagnostics/noresm/$USER

Task 2

Task 2.1

Diagnose only ocean component with passive mode -p

# Example
# (replace with your own experiments)
$ diag_run -m blom -c NHIST_f19_tn14_20190710 -s 1985 -e 2014 -p

In the standard output, you can find lines like:

...
BLOM DIAGNOSTICS SUCCESSFULLY CONFIGURED in /projects/NS2345K/diagnostics/noresm/out/$USER/BLOM_DIAG
...

Go the that directory and check the shell script there, which is job script for each component. You can find the main script for the ocean component blom_diag_template.sh.

Task 2.2

In the configuration file by Task2.2, e.g., by default /projects/NS2345K/diagnostics/noresm/out/$USER/BLOM_DIAG/blom_diag_template.sh

Switch on only some sets, e.g. set_1 and set_3, and switch off other sets
change:

set_1=1
set_3=1

set all other sets_*=0

and then submit the job script, i.e, ./blom_diag_template.sh

Task 2.3

Plot only part of the period of a simulation, instead of the whole period.
Only plot part of the time series between xxx and xxx

Change for example:
/projects/NS2345K/diagnostics/noresm/out/$USER/BLOM_DIAG/blom_diag_template.sh

TRENDS_ALL=0
FIRST_YR_TS1=1
LAST_YR_TS1=10

and then resubmit blom_diag_template.sh

Task 2.4

Diagnose only atmospheric component with passive mode -p, and switch on the chemistry sets and significance test when comparing two models.

Note, it requires at least 10 years to have the significance test for differnce-plots.

# Example
$ diag_run -m cam -c NHIST_f19_tn14_20190710 -s 1985 -e 2014 -p

In the standard output, you can find lines like:

...
CAM DIAGNOSTICS SUCCESSFULLY CONFIGURED in /projects/NS2345K/diagnostics/noresm/out/$USER/CAM_DIAG
...

Go the that directory and check the shell script there, amwg_template.csh

Set in: /projects/NS2345K/diagnostics/noresm/out/$USER/CAM_DIAG/amwg_template.csh

set significance = 0        # (0=ON,1=OFF)
...
set all_waccm_sets = 0 # (0=ON,1=OFF)  Do all the WACCM sets
set all_chem_sets = 0  # (0=ON,1=OFF)   Do all the CHEM sets

then start the scripit ./amwg_template.csh.

Task 3

Task 3.1

Find out where the climo_ts/, config/, logs/, diag/ locate, and understand these processed data e.g., /projects/NS2345K/diagnostics/noresm/out/$USER/BLOM_DIAG

/projects/NS2345K/diagnostics/noresm/out/yanchun/BLOM_DIAG/diag/N1850frc2_f19_tnx1v4_workshop /projects/NS2345K/diagnostics/noresm/out/yanchun/BLOM_DIAG/diag

Task 3.2

Q: How to edit the source code and apply your change?

A: The best way is to fork the NorESM Diagnostic Tool repository, and clone and install your own copy.