Skip to content

Latest commit

 

History

History
197 lines (151 loc) · 10.7 KB

README.md

File metadata and controls

197 lines (151 loc) · 10.7 KB
Raw

GOTM work

A work directory of GOTM, which contains a set of scripts and tools to preprocess the input data, set up runs, and analyze and visualize the output data.

Quick Start

Installation

Create a local copy of gotmwork by

git clone https://github.com/qingli411/gotmwork

Setup

The script setup_gotmwork.sh provides a step-by-step guide to set up the work environment for GOTM. These steps include:

  1. Set up environment variables.
  2. Download source code of CVMix.
  3. Download source code of GOTM.
  4. Compile CVMix.
  5. Compile GOTM.
  6. Set up Python environment for GOTM via Conda.

To use it, chanage directory to gotmwork, type in

./setup_gotmwork.sh

and follow the instructions.

Here is a brief description of each step:

  • Step 1 generates a file .gotmwork_env.sh in the HOME directory which saves all the necessary environment variables. You may optionally add source $HOME/.gotmwork_env.sh in the file $HOME/.bashrc (bash shell) to automatically set up these environment variables when opening a new terminal.
  • Step 2 downloads the source code of CVMix from Github and checkout the required tag. See scripts/get_cvmix.sh.
  • Step 3 downloads the source code of GOTM from Github and checkout the required tag. See scripts/get_gotm.sh.
  • Step 4 compiles CVMix library using build_cvmix.sh. You will be asked to input a Fortran compiler (e.g., gfortran) and the path of netCDF when running it for the first time. See CVMix Homepage for more information on how to build CVMix.
  • Step 5 compiles GOTM using build_gotm.sh. CVMix has to be built before building GOTM.
  • Step 6 sets up Python environment for preprocess data and analysis & visualization using Conda. It will create a new conda environment gotm from gotm_env_*.yml and activate it using set_conda_env.sh.

Compile CVMix and GOTM

In cases of compiling steps are skipped when setting up the gotmwork environment or source code of CVMix and GOTM are changes, CVMix and GOTM can be compiled by

./build_cvmix.sh [-clean] [-build]

and

./build_gotm.sh [-clean] [-build]

Use optional arguments -build to build, -clean to clean the old build, and -clean -build to do a clean build.

Preprocess data

tools/ and scripts/ contain some tools and scripts to preprocess data for GOTM. tools/gotmtool.py provides functions (e.g., write_ts(), write_pfl() and write_spec()) to write data to file in the format that GOTM5 requires.

Run a case

Change directory to a test case (see Test Cases for more detail) and run

./case_run

Analysis and Visualization

tools/gotmanalysis.py provides classes and functions for data analysis and visualization. Some examples using these classes and functions are in visualization/examples.

Test Cases

Test cases. In each case, case_test sets up the namelist, preprocess the input data and run the simulation, whereas case_run (if exists) runs the simulation from preprocessed input data. case_postproc.sh contains steps to postprocess output data and is used by case_run and case_test, either for visulaization or manipulation of the data.

  • COREII Run one set of simulations in each 4 by 4 degree box to cover the global ocean, foced by CORE-II.

    • case_test_multi sets up multiple runs under CORE-II forcing.
    • case_run_multi is similar to case_test_multi, but uses preprocessed CORE-II data and is therefore significantly faster.
    • do_parallel automatically submit parallel jobs to multiple cores.
    • kill_all kills all the jobs.
    • preproc_data preprocesses the CORE-II data.

  • JRA55do Run one set of simulations in each 4 by 4 degree box to cover the global ocean, foced by JRA55-do.

    • case_run_multi sets up multiple runs under JRA55-do forcing using preprocessed data.
    • do_parallel automatically submit parallel jobs to multiple cores.
    • kill_all kills all the jobs.
    • preproc_data preprocesses the JRA55-do data.

  • OCSKEO

  • OCSPapa Single site simulation forced by Ocean Station Papa data.

    • case_postproc.sh is a Bash script to postprocess a single run, used by case_run.

  • OSMOSIS Single site simulation forced by OSMOSIS data.

  • TEST_RES Sensitivity test of different boundary layer schemes to different vertical resolutions and time steps.

    • case_loop.sh loops over different turbulent methods, vertical resolutions and time steps.
    • OCSPapa runs test case using OCS Papa data
    • OSMOSIS runs test case using OSMOSIS data
    • COREII runs test case using selected COREII data

  • Idealized_Tests Set up cases with constant surface forcing and idealzied initial conditions.

  • Idealized_Hurricane Set up the idealized hurricane cases of Reichl et al., 2016.

  • Idealized_Tests_LF17 Set up idealized cases using the initial conditions and surface forcing conditions of Case S-L1 and Case S-B in Li and Fox-Kemper, 2017 (see their Table~1).

  • Idealized_Tests_MSM97 Set up idealized cases using the initial conditions and surface forcing conditions of McWilliams et al., 1997.

Preprocessed Data

The preprocessed input data and namelists for Test Cases are in the directrory data/:

  • OCSPapa_20120101-20131204
  • OSMOSIS_winter
  • OSMOSIS_spring
  • COREII_LAT-54_LON254_20080601-20091231
  • COREII_LAT10_LON86_20080601-20091231
  • COREII_LAT2_LON234_20080601-20091231
  • Idealized
  • Idealized_Hurricane
  • Idealized_Tests_LF17
  • Idealized_Tests_MSM97

In each directory the tool update_nml can be used to update the namelist from data/namelist/ in the case where new entries are added.

Also included in this directory are the data description files in XML format, which are used by case_preproc to preprocess the input data.

Namelist

All namelists are in the directory data/namelist/. Use init_namelist to generate namelist from schemas in the source code according to the type of turbulence closure. It requires Python2 environment and the tool editscenario, which can be installed using the script scripts/install_python_tools.sh.

A List of Tools

A list of tools in the directory tools/. Most of the tools listed below are written in Python3, some in Bash script. The file gotmtool.py contains some shared Python3 functions used by many of the tools. Option -h can be used with all tools to get the usage.

Tool name Description
argo_mld Read Argo temperature and salinity profiles in GOTM input data format and return mixed layer depth based on density threshold
case_preproc Preprocess the input data for GOTM and modify the namelist according to the input xml file.
gotm_archive_data Compress and archive GOTM output data.
gotm_extract_data Extract data from archive generated by gotm_archive_data.
gotm_map_quality_control Remove runs with NaNs in the output data.
is_sea Check if the point given by latitude and longitude is a sea point.
nc2dat Convert observational data (OCS etc.) in netCDF format to formatted text file for GOTM input.
nc2dat_argo Convert Argo profile data in netCDF format to formatted text file for GOTM input.
nc2dat_cdip_spec Convert CDIP wave spectrum data in netCDF format to formatted text file for GOTM input.
nc2dat_core2swr Read the daily maximum shortwave radiation from COREII in netCDF format, add an idealized diurnal cycle, and output the hourly shortwave radiation into formatted text file for GOTM input.
nc2dat_latlon Select CORE-II/JRA55-do/CESM data (in netCDF format) at given latitude and longitude and output into formatted text file for GOTM input.
nc2dat_ww3 Convert WW3 wave variables and partitioned surface Stokes drift data in netCDF format to formatted text file for GOTM input.
nmlchange Change the entry value of a namelist.
nmlquery Query the value of en entry in a namelist.
plotpfl Plot Hovmoller diagram (time-depth) from GOTM output.
plotts Plot time series from GOTM output. Accept multiple variables.

Other Tools

  • gotmanalysis.py contains classes and functions for data analysis and visualization.
  • windwave.py contains tools to estimate and test Stokes drift computed from empirical wind wave spectrum.

A List of Scripts

A list of scripts in the directory scripts/. Bash scripts to setup the tools and runs, and Matlab and NCL scripts to preprocess the input data.

Script name Description
argo_mat2nc.m Matlab script to convert Argo profile data from MAT to netCDF.
case_turbmethod.sh Bash script to set the namelist according to the turbulent methods.
cesm_prep_fluxes.ncl NCL script to prepare the surface fluxes data from CESM output.
cesm_prep_profiles.ncl NCL script to prepare the temperature and salinity profiles from CESM output.
cesm_ww3a_to_gx1v6.ncl NCL script to interpolate the WW3 output data onto POP grid gx1v6.
core2_prep_meteo.ncl NCL script to prepare meteorology data from CORE-II.
get_cvmix.sh Bash script to download CVMix source code from Github.
get_gotm.sh Bash script to download GOTM source code from Github.
install_python_tools.sh Bash script to download and install Python tools for GOTM from Github, including editscenario, xmlstore, xmlplot and gotmgui.
jar55do_prep_meteo.ncl NCL script to prepare meteorology data from JRA55-do.
ocs_heatflux.ncl NCL script to prepare the net heat flux (excluding shortwave) data from longwave, sensible and latent heat fluxes for GOTM.
roms_dz.m Matlab script to generate ROMS style stretching vertical grid.
set_gotmwork_env.sh Bash script to set up environment variables for GOTM.

Other Scripts

Scripts in the root directory:

  • build_cvmix.sh is a Bash script to build GOTM.
  • build_gotm.sh is a Bash script to build GOTM.
  • set_tools.sh is a Bash script to set up paths and tools, used by case_run.
  • setup_gotmwork.sh is a Bash script to set up work environment for GOTM.