2. NorESM2 Diagnostics Package

2.1. Introduction

The NorESM Diagnostic Package is a NorESM model evaluation tool written with a set of scripts and command utilities (bash/cshrc, NCO, CDO and NCL etc) to provide a general evaluation and quick preview of the model performance with only one command line. This tool package works on the original model output and has NorESM-specific diagnostics.

The tool package consists of:

CAM_DIAG

The CAM_DIAG diagnostics package computes climatological means of the simulation(s) and produces plots and tables of the mean climate in the atmospheric model component CAM-Nor. Based on NCAR’s AMWG Diagnostics Package

CLM_DIAG

The CLM_DIAG diagnostics package computes climatological means of the simulation(s) and produces plots and tables of the mean climate in the land model component CLM. Based on CLM Land Model Diagnostics Package

CICE_DIAG

The CICE_DIAG diagnostics package computes climatological means of the simulation(s) and produces timeseries of e.g. snow/sea ice volume/area in the sea ice model component CICE.

HAMOCC_DIAG

The HAMOCC_DIAG diagnostics package computes and produces time series, climaotology, zonal and regional averages of varaibles from the biogeochemistry model component iHAMOCC.

BLOM_DIAG

The BLOM_DIAG diagnostics package computes and produces climaotology, time series, zonal and regional averages, volume transports and fluxes in the ocean model component BLOM.

(See more on the Major changes to the NCAR’s Diagnostics Package at the bottom)

2.2. Installation

The source codes of the NorESM diagnostics packages are developed and maintained on the Github: https://github.com/NordicESMhub/noresmdiagnostics.

And the observation dataset and grid files are hosted at: https://www.noresm.org/diagnostics, with a total size of ~100 GB.

One can either use the preinstalled package or has a fresh install, as described below:

2.2.1. Preinstalled package

The full diagnostic package (including source files and data files) are currently installed on storage platform NIRD and the HPC computing Betzy. It is not supported yet on Fram.

2.2.1.1. On NIRD

The pre-installed path:

NIRD:/projects/NS2345K/diagnostics/noresm

You don’t need to install this diagnostic package, but you can call it as a command line directly on NIRD. As a prerequiste, you should have access permission to the NS2345K project on NIRD.

One can either add the diag_run to your search path:

export PATH=$PATH:/projects/NS2345K/diagnostics/noresm/bin

(assuming you are using Bash Shell)

or add it as an alias in $HOME/.bashrc:

alias diag_run="/projects/NS2345K/diagnostics/noresm/bin/diag_run"

and source ~/.bashrc to make these changes take effect.

2.2.1.2. On Betzy

The pre-installed path:

Betzy:/cluster/shared/noresm/diagnostics/noresm

You can run this package without installation as a noresm user group. There are two wrapper scripts:

  • diag_run for running on the Betzy login nodes (highly discouraged, see Sigma2 HPC policy)

  • diag_srun to submit a SBATCH job to the compute node of Betzy as a preproc type of job.

One can either add the diag_run and diag_srun to your search path:

export PATH=$PATH:/projects/NS2345K/diagnostics/noresm/bin

(assuming you are using Bash Shell)

or add it as an alias in $HOME/.bashrc:

alias diag_run="/projects/NS2345K/diagnostics/noresm/bin/diag_run"
alias diag_srun="/projects/NS2345K/diagnostics/noresm/bin/diag_srun"

and source ~/.bashrc to make these changes take effect.

For details on how to invoke the diag_run or diag_srun on Betzy, please refer to the subsection: Run the tool on Betzy.

Note

DO NOT make changes direclty in this preinstalled package; if you are willing to modify the code, refer to the next subsection below.

2.2.2. Fresh install

If you want to change the code for your own purpose, you can installed it on NIRD under your HOME folder or your own project area (i.e., /projects/NSxxxxK). There are several steps to install it:

  1. Fork the NorESM Diagnostic Package Github repository to your own Github respository, e.g., https://github.com/YOU_GITHUB_USERNAME/noresmdiagnostics

  2. Change to your preferred location, denoted as $DIAGROOT, where you want to install the tool, and git clone https://github.com/YOU_GITHUB_USERNAME/noresmdiagnostics

  3. Change to $DIAGROOT/noresmdiagnostics/bin, and link or download all the observation and grid data files.

  • If you are installing the tool on NIRD, you just need to link all the data to your clone by running the script linkdata.sh, given you have access to the /project/NS2345K project

  • If you are not memember of NS2345K or you are installing it on platforms other than NIRD, you should download all the data to your clone by executing dloaddata.sh. If you are not running it on NIRD, you should have CDO, NCO and NCL installed.

  1. Make changes to the code/scripts for your purpose. And call diag_run of your own clone.

  2. If you would like to contribute your function enhancements or bug fixes to the original diagnostic package, you should commit the changes to your fork repository, then create an Issue at the Github repository, and finally make a pull request to the original Github repository to incorporate your changes.

2.3. Run the tool on NIRD

On NIRD, each package can be run/configured from the command line using the wrapper script for NorESM diagnostic program diag_run.

On Betzy, one can run the tool directly on the login nodes (not recommended), or by submitting a preproc job to the compute nodes, please refer to the subsection for more Betzy-specific details: Run the tool on Betzy.

Call the wrapper script with diag_run -h will give you the description of the command-line options:

-------------------------------------------------
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

Command-line options:
-m, --model=MODEL                             Specify the diagnostics package (REQUIRED).
                                              Valid arguments:
                                                cam    : atmospheric package (AMWG)
                                                clm    : land package (LMWG)
                                                cice   : sea-ice package
                                                blom   : ocean package
                                                hamocc : biogeochemistry package
                                                all    : configure all available packages.
-c, -c1, --case=CASE1, --case1=CASE1          Test case simulation (OPTIONAL).
-s, -s1, --start_yr=SYR1, --start_yr1=SYR1    Start year of test case climatology (OPTIONAL).
-e, -e1, --end_yr=EYR1, --end_yr1=EYR1        End year of test case climatology (OPTIONAL).
-c2, --case2=CASE2                            Control case simulation (OPTIONAL).
-s2, --start_yr2=SYR2                         Start year of control case climatology (OPTIONAL).
-e2, --end_yr2=EYR2                           End year of control case climatology (OPTIONAL).
-i, -i1, --input-dir=DIR, --input-dir1=DIR    Specify the directory where the test case history files are located (OPTIONAL).
                                              Default is --input-dir=/projects/NS2345K/noresm/cases
-i2, --input-dir2=DIR                         Specify the directory where the control case history files are located (OPTIONAL).
                                              Default is --input-dir=/projects/NS2345K/noresm/cases
-o, --output-dir=DIR                          Specify the directory where the package(s) the climatology and time-series files should be stored (OPTIONAL).
                                              Default is --output-dir=/projects/NS2345K/diagnostics/noresm/out/$USER
-p, --passive-mode                            Run the script in passive mode: the diagnostic script will be configured but not executed (OPTIONAL).
-t, --type=TYPE                               Specify climatology or time series diagnostics (OPTIONAL): valid options are --type=climo and --type=time_series.
                                              Default is to run both. Note that the time series are computed over the entire simulation.
-w, --web-dir=DIR                             Specify the directory where the html should be published (OPTIONAL).
                                              Default is --web-dir=/projects/NS2345K/www/diagnostics/noresm/$USER
--no-atm                                      Run CLM diagnostics without CAM data. Must be used for offline CLM simulations.

Examples:
diag_run -m all -c N1850_f19_tn11_exp1 -s 21 -e 50 # model-obs diagnostics of case=N1850_f19_tn11_exp1 (climatology between yrs 21 and 50) for all model components.
diag_run -m cam -c N1850_f19_tn11_exp1 -s 21 -e 50 -w /path/to/my/html # model-obs diagnostics in CAM, publish the html in /path/to/my/html.
diag_run -m blom -c N1850_f19_tn11_exp1 -t time_series # model-obs time-series diagnostics in BLOM for all years represented in the model output directory (/projects/NS2345K/noresm/cases/N1850_f19_tn11_exp1/ocn/hist/).
diag_run -m cice -c N1850_f19_tn11_exp1 -s 21 -e 50 -p # configure (but do not run) model-obs diagnostics for CICE.
diag_run -m clm -c N1850_f19_tn11_exp1 -s 21 -e 50 -i /input/directory1 -c2 N1850_f19_tn11_exp2 -s2 21 -e2 50 -i2 /input/directory2 # model1-model2 diagnostics for CLM with user-specified history file directories
diag_run -m blom -c N1850_f19_tn11_exp1 -s 21 -e 50 -t climo # model-obs climatology diagnostics (no time series) for BLOM:
diag_run -m cam -o /my/dir # install CAM diagnostics in /my/dir with minimal configuration.
diag_run -m blom,hamocc -c N1850OC_f19_tn11_exp1 -s 21 -e 50 # model-obs diagnostics for BLOM and HAMOCC.
diag_run -m clm -c N1850_f19_tn11_clmexp1 -s 71 -e 100 --no-atm # model-obs time-series diagnostics for an offline (uncoupled) CLM simulation.
diag_run -m hamocc -c N1850OC_f19_tn11_exp1 -s 31 -e 100 -t time_series # model-obs time-series diagnostics in HAMOCC between yrs 31 and 100.

*** NOTE: '-m micom' should be used for the ocean component of NorESM version 1 ***

2.3.1. Description

diag_run is a wrapper script, which is used to run the diagnostics for each NorESM component (cam, clm, cice, blom, and hamocc). The diagnostic packages can be used to plot model results with respect to either observations (so-called model-obs diagnostics), or to another simulation (model1-model2 diagnostics). The diagnostics for the atmosphere (cam), land (clm) and sea-ice (cice) are based on the NCAR packages, but has undergone some major improvements, particularly in the climatology and time-series computations. The ocean (blom) and its biogeochemistry (hamocc) have been developed in-house.

Please note, the ocean component of the NorESM2, BLOM, is an updated version of MICOM. It is named MICOM in NorESM1 for CMIP5 experiments and in NorESM2 for many (but not all) CMIP6 experiments. Therefore, for experiments with MICOM as the ocean component of NorESM, -m micom can be used in the command line option for diag_run. To compare a simulation with either MICOM or BLOM to the other (model1-model2 diagnostics), you can either use -m micom or -m blom. Both options should work.

diag_run has two modes:

  • active-mode, for which diag_run runs the diagnostic scripts

  • passive-mode, for which diag_run only configures the scripts.

In the passive-mode the diagnostic scripts have to be run manually by the user. By default, diag_run is always in the active-mode, but switches into passive-mode if at least one of these two criteria are fulfilled:

  1. The user invokes the option -p (see below), or

  2. The user does not give enough information needed to run the diagnostics (next subsection).

2.3.2. Active-mode

If you want to use diag_run to run the full (climatology and time-series) diagnostics, the minimum requirement is to specify the options of model, case_name, start_yr and end_yr (-m, -c, -s and -e).

Example 1:

diag_run -m cam -c N1850_f19_tn14_191017 -s 21 -e 50

This command runs atmospheric model-obs diagnostics of the case N1850_f19_tn14_191017 using a climatology between model years 21 and 50. It is assumed that the N1850_f19_tn14_191017 history files are located under /projects/NS2345K/noresm/cases. By default, the resulting plots and html will be stored in

/projects/NS2345K/www/diagnostics/noresm/<username>/N1850_f19_tn14_191017/CAM_DIAG

where <user_name> is your NIRD username. Or if you specify to store them under a common folder, i.e. with -w /projects/NS2345K/www/diagnostics/noresm/common. It links to the following URL: https://ns2345k.web.sigma2.no/diagnostics/noresm/common/N1850_f19_tn14_191017/CAM_DIAG/yrs21to50-obs/sets.htm.

The climatology and time-series files under

/projects/NS2345K/diagnostics/noresm/out/<username>/CAM_DIAG

If you want to run model1-model2 diagnostics, you also need to specify case_name2, start_yr2 and end_yr2 (-c2, -s2, -e2) in addition.

Example 2:

diag_run -m cam -c N1850_f19_tn14_191017 -s 21 -e 50 -c2 B1850MICOM_f09_tn14_01 -s2 21 -e2 50

would be the same as in Example 1 above, except for comparing N1850_f19_tn14_191017 to B1850MICOM_f09_tn14_01 instead of observations.

In Example 1 and Example 2 the options -s and -e (as well as -s2, -e2) refer to the start and end years of the climatology. The time-series are calculated from all the history files in the case directory (input_dir). This is always the case unless the user invokes the option -t time_series. If this option is invoked, start_yr and end_yr refer to the beginning and end of the time series instead of the climatology, hence:

Example 3:

diag_run -m blom -c N1850_f19_tn14_blom_20200608 -t time_series -s 1 -e 10

would produce blom time-series plots between years 1 and 20. Note that omitting start_yr and end_yr when the option -t time_series is invoked computes the time-series over the entire experiment (all history files in the case directory, input_dir):

diag_run -m cam -c N1850_f19_tn14_191017 -t time_series

diag_run uses some template scripts for each of the model components. When diag_run is executed, these scripts are changed according to the user-specified settings and renamed with a time stamp. For example, if you run the blom diagnostics, the run script template (blom_diag_template.sh) will be renamed with a time-stamp as blom_diag_YYMMDD_HHMMSS.

diag_run also creates a config and output file with the same time stamp (config_YYMMDD_HHMMSS and out_YYMMDD_HHMMSS, respectively). The config file stores information about changes in the diagnostics scripts invoked by the user, and the output file contains the standard output and error (i.e. what is shown in your terminal during runtime). When the diagnostics a component is finished the run scripts are copied to:

output_dir/<user_name>/XXX_DIAG/config/<casename>/run_scripts

and the config and output files to:

output_dir/<username>/XXX_DIAG/config/case_name/logs

Hence, for Example 1 above, the run scripts are saved in:

/projects/NS2345K/diagnostics/noresm/out/<username>/CAM_DIAG/config/N1850_f19_tn14_191017/run_scripts

and the config and out files in:

/projects/NS2345K/diagnostics/noresm/out/<username>/CAM_DIAG/config/N1850_f19_tn14_191017/logs

2.3.3. Passive-mode

Another important property of diag_run is that it will only run the diagnostics if sufficient information has been provided by the user; otherwise it switches into passive-mode. diag_run will then configure the diagnostics scripts as much as possible (based on the information provided by the user), and also add information to the config file about which variables are still required to be modified by the user in order to run the diagnostic script. This option is particularly useful if you want to do some development work on the diagnostics scripts, or if you want to change any variables in the diagnostics scripts that are not included as an option in diag_run. Hence, if you run the following command:

diag_run -m clm

the following will appear on the screen:

[nird@login0 ~]$ /projects/NS2345K/diagnostics/noresm/bin/diag_run -m clm
-------------------------------------------------
Program:
/projects/NS2345K/diagnostics/noresm/bin/diag_run
Version: 2.1
-------------------------------------------------
-CHANGING DIAGNOSTICS DIRECTORY to /projects/NS2345K/diagnostics/noresm/out/<username>/CLM_DIAG in lnd_template.csh
-CHANGING ROOT DIRECTORY FOR CODE AND DATA to /projects/NS2345K/diagnostics/noresm/packages/CLM_DIAG in lnd_template.csh
-CHANGING INPUT DIR 1 to /projects/NS2345K/noresm/cases in lnd_template.csh
-CHANGING publish_html_root to /projects/NS2345K/www/diagnostics/noresm/<username> in lnd_template.csh
-SETTING UP TIME-SERIES DIAGNOSTICS FOR ENTIRE EXPERIMENT
CLM DIAGNOSTICS SUCCESSFULLY CONFIGURED in /projects/NS2345K/diagnostics/noresm/out/<username>/CLM_DIAG
-------------------------------------------------
lnd_template.csh IS NOT RUNNING: NOT ALL REQUIRED VARIABLES HAVE BEEN CONFIGURED (see /projects/NS2345K/diagnostics/noresm/out/<username>/CLM_DIAG/config.log).
-------------------------------------------------
-------------------------------------------------
TOTAL diag_run RUNTIME: 0m1s
-CLM diagnostics: 0m1s
-------------------------------------------------
DONE: Tue Dec 22 12:47:49 CET 2020

The (semi-configured) run script has then been copied to /projects/NS2345K/diagnostics/noresm/out/<username>/CLM_DIAG/lnd_template.csh, and all information about the configuration is contained in /projects/NS2345K/diagnostics/noresm/out/<username>/CLM_DIAG/config.log

2.3.4. Options

diag_run options (flags) typically come in both short (single-letter) and long forms. A complete description of all options is given below in alphabetical order of the short option letter. When invoked without options, diag_run prints a table containing all options along with some examples (see also below).

-c case_name (-c1, --case, --case1)

Name of the test case experiment that you want to run diagnostics for. This option is required if you want to use diag_run in active-mode.

-c2 case_name2 (--case2)

Name of the control case experiment. This option is required if you want to run model1-model2 diagnostics in active-mode.

-e end_year (-e1,--end_yr,--end_yr1)

If --type=time_series, this option refers to the end year of time-series for case_name. Otherwise, it refers to the end year of climatology. This option is optional if --type=time_series, but required for active-mode diagnostics if --type=climo or if type is not invoked.

-e2 end_year (--end_yr2)

If --type=time_series, this option refers to the end year of time-series for case_name2. Otherwise, it refers to the end year of climatology. This option is optional if --type=time_series, but required for active-mode model1-model2 diagnostics if --type=climo or if type is not invoked.

-i input_dir (-i1, --input-dir, --input-dir1)

Name of the root directory of the monthly history files for case_name. For example, if your blom history files are located in /this/is/a/directory/case1/ocn/hist, this option should be set to input_dir=/this/is/a/directory. Default is input_dir=/projects/NS2345K/noresm/cases .

-i2 input-dir2 (--input-dir2)

Name of the root directory of the monthly history files for case_name2. Also here, default is input_dir2=/projects/NS2345K/noresm/cases .

-m model (--model)

Name of the model you want to run the diagnostics for. Valid options are cam, clm, cice, blom, hamocc and all. This is the only option that is required for both the active and passive mode. If you invoke the “all” option, the cam, clm, cice, blom and hamocc diagnostics will be run subsequently. It is also possible to combine different models as you wish within this option: for example, if you only want to run cam and clm diagnostics, you can simply add the names of those models and separate them with a comma (-m cam,clm).

--no-atm

This option, which takes no argument, skips the usage of CAM history files in the CLM diagnostics. This option is necessary for offline CLM simulations.

-o output_dir (--output_dir)

Root directory where you want to store the output from the diagnostics (i.e. the climatology and time-series files). For example, if you set output_dir=/just/another/directory, the climatology and time-series files from the blom diagnostics will be stored in:

/just/another/directory/BLOM_DIAG/.

Default is:

output_dir=/projects/NS2345K/diagnostics/noresm/out/<username>

where <username> is your user name on NIRD.

-p, --passive-mode

This option, which takes no argument, forces diag_run into passive-mode. This means, even if you have given sufficient information to run in active-mode, the diagnostic scripts will not be executed.

-s start_year (-s1,--start_yr,--start_yr1)

If --type=time_series, this option refers to the start year of time-series for case_name. Otherwise, it refers to the start year of climatology. This option is optional if --type=time_series, but required for active-mode diagnostics if --type=climo or if type is not invoked.

-s2 start_year2 (--start_yr2)

If --type=time_series, this option refers to the start year of time-series for case_name2. Otherwise, it refers to the start year of climatology. This option is optional if --type=time_series, but required for active-mode model1-model2 diagnostics if --type=climo or if type is not invoked.

-t type (--type)

Specifies if you only run climatology or time-series diagnostics: valid options are --type=climo and –type=time_series. Default is to run both.

-w webdir (--web-dir)

Specifies the directory where the html should be stored. This directory should preferably be linked to a web server so that one can look at the results with a web browser. Default is:

--web-dir=/projects/NS2345K/www/diagnostics/noresm/

2.3.5. Examples

Model-obs diagnostics of case=N1850_f19_tn11_exp1 (climatology between yrs 21 and 50) for all model components:

diag_run -m all -c N1850_f19_tn11_exp1 -s 21 -e 50

Model-obs diagnostics in CAM, publish the html in /path/to/my/html:

diag_run -m cam -c N1850_f19_tn11_exp1 -s 21 -e 50 -w /path/to/my/html

Model-obs time-series diagnostics in BLOM for all years the model output directory (/projects/NS2345K/noresm/cases/N1850_f19_tn14_blom_20200608/ocn/hist/):

diag_run -m blom -c N1850_f19_tn14_blom_20200608 -t time_series

Configure (but do not run) model-obs diagnostics for CICE:

diag_run -m cice -c N1850_f19_tn11_exp1 -s 21 -e 50 -p

Model1-model2 diagnostics for CLM with user-specified history file directories:

diag_run -m clm -c N1850_f19_tn11_exp1 -s 21 -e 50 -i /input/directory1 \
-c2 N1850_f19_tn11_exp2 -s2 21 -e2 50 -i2 /input/directory2

Model-obs climatology diagnostics (no time series) for BLOM:

diag_run -m blom -c N1850_f19_tn14_blom_20200608 -s 1 -e 10 -t climo

Install CAM diagnostics in /my/dir with minimal configuration:

diag_run -m cam -o /my/dir

Model-obs diagnostics for BLOM and HAMOCC:

diag_run -m blom,hamocc -c N1850_f19_tn14_blom_20200608 -s 1 -e 10

Model-obs time-series diagnostics for an offline (uncoupled) CLM simulation:

diag_run -m clm -c N1850_f19_tn11_clmexp1 -s 71 -e 100 --no-atm

Model-obs time-series diagnostics in HAMOCC between yrs 31 and 100:

diag_run -m hamocc -c N1850OC_f19_tn11_exp1 -s 31 -e 100 -t time_series

2.4. Run the tool on Betzy

There are two alternatives to run the tool on Betzy, either as an interactive (for short test and debug runs) or a batch job (recommended). It is also possible to run directly on the login node with diag_run, but it is higly discouraged and not an option (Refer to Sigma2 HPC policy).

The main purpose to run the tool on Betzy is to get a quick diagnostic of model output when the model is still on-the-fly, but already has some intermediate output been short-term archived to /cluster/work/users/<username>/archive (Refer to Archiving NorESM output).

Since the mounted NIRD project disks /nird/projects/NSxxxxK are not accessible from the compute nodes, the -i, -o have to point to /cluster/work/users/<username>/<workdir>, with an execption for the -w option. See explanations and examples in the following.

2.4.1. As interactive job

Run with an intactive sbatch job, with diag_run.

Start an interactive job request by, for example :

$ salloc --nodes=1 --mem-per-cpu=12G --time=00:30:00 --partition=preproc --account=nn2345k

And then use the same command-line options of diag_run as on NIRD.

2.4.2. As batcth job

Submit a backend preproc job with diag_srun.

There are sbatch job specific command-line options for diag_srun, in addition to the diag_run -h options:

--account=nsxxxxk                             : (OPTIONAL. Project account for CPU hours (default ns2345k).
--time=DD-HH:MM:SS                            : (OPTIONAL. CPU walltime (default value according to length of years and active components).
--remove-source-files-flag=true|false         : (OPTIONAL. Flag if the source file will be removed after the webpage is moved from Betzy /cluster to NIRD /project* (default as false).

See diag_srun -h for the help information.

Examples:

  1. Use all default settings

    $ ./diag_srun -m blom -c test_case_name -s 1 -e 10

It is the same as diag_run.

  1. Set CPU account and hours

    $ ./diag_srun -m blom -c test_case_name -s 1 -e 10 --account=nn2345k --time=0-00:59:00

The CPU account is set to default as nn2345k if not prescribed. The CPU hours is set according the prescribed experiment start and end years.

  1. Set input data, output data, and webpage path to /cluster on Betzy (the same as default values)

    $ ./diag_srun -m blom -c NOICPLHISTOC_f09_tn14_cpldiags -s 1 -e 20 -i /cluster/work/users/<username>/archive -o /cluster/work/users/<username>/diagnostics/out -w /cluster/work/users/<username>/diagnostics/www

The above settings for -i, -o and -w are default values if they are not prescribed. As the mounted NIRD disks /nird/projects are not accessible from the compute nodes, the -i and -o options have to be set to /cluster**. For the -w option, see the next example.

  1. Set input data, output data on Betzy, and webpage path on NIRD

    $ ./diag_srun -m blom -c NOICPLHISTOC_f09_tn14_cpldiags -s 1 -e 20 -w /nird/projects/NS2345K/www/diagnostics/noresm/<username> --account=nn2345k --time=0-00:59:00

The created webpage will saved to NIRD. The webpage path specificed by -w will temporary set to the defaut location under /cluster/work/users/<username>/diagnostics/www, and will rsync to NIRD after the diagnostics job is finished.

  1. Remove source webpage files from Betzy after transferred to NIRD

    $ ./diag_srun -m blom -c NOICPLHISTOC_f09_tn14_cpldiags -s 1 -e 20 -w /nird/projects/NS2345K/www/diagnostics/noresm/<username>/ --remove-source-files-flag=true

Options to set if temporary webpage under /cluster as described above will be removed after they are transferred to NIRD (only valid if -w option is set to /nird/projects area)

See more help:

/cluster/shared/noresm/diagnostics/noresm/bin/diag_run -h
/cluster/shared/noresm/diagnostics/noresm/bin/diag_srun -h

Note

The mounted NIRD project area /nird/projects are not available on the HPC computing nodes. Therefore, the -i, -o can only be set to locations under /cluster/work/users/<username>. The -w option can be set to /nird/projects area to facility the browsing the webpage-based diagnostics. It is actually set to /cluster during runtime, but transfer the created webpages to NIRD automatically after the diagnostic is finished.

Major changes to the NCAR’s Diagnostics Package

The diagnostic tool package is based on NCAR’s CAM and CLM Diagnostic Packages.

  • Changes to all components

    The following major changes have been made in all diagnostic packages:

    • The calculation of the climatology has been improved, using the ncclimo oporator from nco.

    • The bash/csh variables publish_html and publish_html_root have been added in order to enable publication of the html on the NIRD web server.

    • There is now the option to calculate time series over the entire simulation (default). Hence, the start and end years of the time series must no longer be specified.

    • The bash/csh variable CLIMO_TIME_SERIES_SWITCH has been added in order to allow for diag_run to compute only climatology or time series if desired.

    • The environmental variable ncclimo_dir has been added in order to allow for diag_run to be run by cron.

  • CAM_DIAG specific major changes

    • The CAM diagnostics (amwg) now calculate the annual and global mean time series of the net TOA radiation balance. The results are published on the web server together with the other figures.

  • CLM_DIAG specific major changes

    • The amount of variables used in the time series calculations have been dramatically reduced in order to reduce time and computational resources

    • If time series or climatology is computed is now determined by the selected sets in the computation.

  • CICE_DIAG specific major changes

    • The switch CNTL has been added in order to determine whether one or two cases should be plotted.

  • BLOM_DIAG (newly developed)

    Two modes of diagnostics: compare to the observations and anothor model run; includes diagnostics of:

    • Time series plots

      1. Sections transports

      2. Global averages

      3. Maximum AMOC

      4. Hovmoeller plots

      5. ENSO indices

    • Climatology plots

      1. Horizontal fields - annual means

      2. Horizontal fields - seasonal/monthly means

      3. Overturning circulation

      4. Zonal means (lat-depth)

      5. Equatorial cross sections

      6. Meridional fluxes (vertically integrated)

  • HAMOCC_DIAG (newly developed)

    Two modes of diagnostics: compare to the observations and another model run; includes diagnostics of:

    • Time series plots

      1. Global fluxes

      2. Global averages

    • Climatology plots

      1. Horizontal fields

      2. Zonal mean fields

      3. Regionally-averaged monthly climatologies

  • CISM_DIAG (newly developed)

    Two modes of diagnostics: compare to the observations and another model run; includes diagnostics of:

    • Time series plots

      1. Mass/ice fluxes

      2. Mass/ice/temperatue averages