NOAA

Geophysical Fluid
Dynamics Laboratory

Skip to: [content] [navigation]
If you are using Navigator 4.x or Internet Explorer 4.x or Omni Web 4.x , this site will not render correctly!

gfdl's home page > people > John Dunne >

Adding SiO4 to the MOM4 OCMIP2_BIOTIC package

Description

This HOWTO describes in detail the process of adding a tracer to MOM4, starting with ocmip2_biotic, assuming that you are already versed in running existing tracers in MOM4. Silicate is used as an example case. The basic idea is to:

1) Copy an existing module to a new name
2) Go through the code looking for lines dealing with phosphate, copying them directly below, and changing them to silicate
3) Add the new module to ocean_tpm.F90
4) create new input files
5) Point the runscript to the new forcing, module and input files


1) Copy an existing module to a new name

Assuming that you have gone through the HOWTO for running the ocmip2_biotic code in MOM4, go to the directory in which the tracer code resides, copy the existing ocmip2_biotic.F90 module to a new name, and open this file in a text editor:

First, browse through this file, noticing that it contains a list of the external modules that it uses and then the subroutines contained within it:
and:

The first edit is rename all references to the module to a new name by doing a global "search" and "replace" to change all 'ocmip2_biotic' references to 'ocmip2_biotic_si'... Thus, the easypart ends, and the copy - paste - change names part begins.

2) go through the code looking for lines dealing with phosphate, copying them directly below, and changing them to silicate.

Edit biotic_type to add:

Edit Private variables to add:

Edit allocate_arrays to add:

Edit ocmip2_biotic_si_end to add:

Edit ocmip2_biotic_si_sbc to add:

Edit ocmip2_biotic_si_set_up to add:

Edit ocmip2_biotic_si_source to add:

Edit ocmip2_biotic_si_start to add:

Still with ocmip2_biotic_si_start, add the following variables to the ocmip2_biotic_glob_nml namelist: Still with ocmip2_biotic_si_start, add the following variables to the ocmip2_biotic_nml namelist: Lower down in ocmip2_biotic_si_start, add: Edit ocmip2_biotic_si_tracer to add: 3) Add the new module to ocean_tpm.F90

Edit ~/jakarta/ocmip2_biotic/src/mom4/ocean_core/ocean_tpm.F90 to add the appropriate lines for ocmip2_biotic_si corresponding to existing lines for ocmip2_biotic:

4) Create new input file(s)

MOM4 must be feed initial and boundary conditions on the model grid in COARDS-compliant netCDF format. For a detailed explanation of preprocessing tools available to create these files, see the MOM4 guide, or follow the Preprocessing guidelines for creation of silicate boundary condition. Briefly, some helpful commands are given below:

Make an input_fields directory in your path (depending on the size of the files, you may need to put this in your /archive... /home is limited to 10 GBytes).

In order to regrid a data field, the preprocessing routine must be fed an input field that has values at all points contained within the model grid - i.e. extrapolated over land and spanning the X, Y, and Z grids. To extrapolate a field in ferret, qlogin to the analysis cluster (AC) and follow these ferret commands: Grid Generation and re-gridding c-shell scripts have been developed by Zhi Liang and are documented in the MOM4 guide. This software is automatically included in the source code of the model, and can be found in:

At the top of regrid_3d.csh, you will need to edit the path/file information in the following lines to suit your particular needs:

In addition, lower down in regrid_3d.csh, you will need to change the variable name information:


In the AC, run the shell script simply by:
If you need to further edit the netCDF file in order to conform to formatting requirements, you may want to edit the file. To do this, you must first dump the file to ascii format and then probably (if it's big) separate the file into parts after finding out how many lines there are:
Make changes to head to conform to formatting... see example in /archive/jpd/postinchon/input_fields/woa01_sio4_om1_bc.nc, which has the following header information (upon ncdump -h [FILENEME]:
Make the necessary edits, then recombine the head and tail and convert back to NetCDF with: 5) Point the runscript to the new forcing, module and input files

Create a new experiment in your jakarta_tracers.xml file called om1_ocmip2_biotic_si, and change all of the tracer-related information:

Make the new ocean_tpm.F90 and ocmip2_biotic_si.F90 files part of the suite of files copied into the source directory tree and linked with the Makefile by adding the lines: Make an input directory in your path and add the new diagnostic fields as optional outputs to a new file: The format of these fields is: Add initial and forcing fields to the list of files used in the run: Make new files corresponding to the inputs for namelist options, initail and boundary conditions, packages to be turned on, etc. including: ocmip2_biotic_si_glob.nml, ocmip2_biotic_si.nml, ocean_tracer_tree_ob_si, ocean_tracer_tree_init_ob_si, ocmip2_biotic_si_res.nml, and ocean_tracer_tree_init_ob_si_res and change the references to them in:
Last modified: Friday, 18-Sep-03 .   Send comments/suggestions to John Dunne (jpd)


smaller bigger reset
last modified:May 04 2004.