3DVAR
Tutorial
e. 3DVAR – Run Case Study
(06/09/2003)
In this section,
you will learn how to run 3DVAR using observations and a first guess from an
actual operational system.
By this stage you
have successfully created the three input files (first guess fg,
observation ob and background error statistics be in directory $DAT_DIR)
required to run 3DVAR. Also, you have successfully downloaded and compiled the
3DVAR code. If this is correct, we are ready to learn how to run 3DVAR. If not,
then you’ll need to return to the 3DVAR
Tutorial Page to complete stages a. to c.
To view the MM5
format analysis output of 3DVAR, we make use of the MM5 GRAPH plotting package.
You will need to download this and compile. To learn how to do this, visit the
MM5 tutorial’s GRAPH
tutorial and proceed up to the point at which you have compiled GRAPH in
directory $DAT_DIR.
1. The Case
The data for this
case comes from the Taiwanese Civil Aeronautics Administration (CAA)’s
MM5-based real-time AOAWS. All available observations valid within 90 minutes
of 12 UTC 19 August 1999 are available. Although the AOAWS system is a
quadruple (135/45/15/5km resolution) two-way nesting system, in this tutorial
we run 3DVAR on only the lowest resolution domain 1. The only reason for this
choice is quick turnaround – domain 1’s dimensions are the smallest of the four
(67x81x31 grid points).
2. Required
3DVAR namelist changes
You will run
3DVAR using the script DA_Run_3DVAR.csh found in subdirectory 3dvar/run.
Open this script with your favorite editor, and spend some time browsing it to
accustom yourself with the layout. Don’t despair, in this tutorial we will not
be changing too much in here. See c.
Setting Up 3DVAR for more information on the various namelist options.
In your first
experiment, the only changes you need to make to the DA_Run_3DVAR.csh
script are to specify the DAT_DIR and RUN_DIR environment variables.
Optionally, you can also change the RUN_ID variable. The script will then
create a new subdirectory ${RUN_DIR}/${RUN_ID} in which 3DVAR will run - all
output is directed there.
3. Run
3DVAR
Once you have set
the necessary environment variables, run 3DVAR for this case study by typing
DA_Run_3DVAR.csh
in the 3dvar/run
subdirectory.
Successful
completion of the job results in a number of output diagnostics files in the
${RUN_DIR}/${RUN_ID} subdirectory. Tthe various textual diagnostics output
files wil be explained in the next section (3DVAR
– Diagnostics). Here, we merely wish to view the analysis increments output
by 3DVAR for this case study.
4. View
3DVAR Analysis Increments
If you have not
already done so, download and compile GRAPH.
As mentioned above,
PSOTs are useful for illustrating the impact of background error covariances in
3DVAR. As an example, first cd to the GRAPH directory. Copy the g_plots.tbl
file from ftp://ftp.ucar.edu/mesouser/mm53dvar/g_plots.tbl
to your GRAPH directory (you may want to increase the contour interval of the g_plots.tbl
file). Then, to run GRAPH, type
graph.csh 1 1
${RUN_DIR}/${RUN_ID}/DAProg_3DVAR.analincs
where
${RUN_DIR}/${RUN_ID} is the directory containing your output. On successful
completion, a file “gmeta” will be produced containing plots of the 3DVAR
analysis increments. This can be viewed by typing
idt gmeta
The file contains
plots of temperature and pressure increments resulting from the assimilation of
all observations. Here is an example vertical cross-section that should be
produced by in your gmeta file.
In order to give
yourself more experience with some of the 3DVAR options, try rerunning 3DVAR
with different namelist.3dvar options. For example,
·
Withholding particular
observation types by setting e.g. DA_USE_SYNOPOBS = .FALSE. (ignores surface
SYNOP observations).
·
Making 3DVAR’s convergence
more stringent by reducing the value of DA_EPS0 to e.g. 0.001. If 3DVAR hasn’t
converged by the maximum number of iterations (DA_NTMAX) then you may need to
increase DA_NTMAX.
As before, you
may wish to change RUN_ID if you want to save all the outputs.
5. What
next?
Having run 3DVAR
and taken at cursory glance at the analysis increments, you should now spend
time looking at some of the diagnostic output files created by 3DVAR:
Miscellanies:
Trouble Shooting: If you have questions, ask mesouser.