Warning, /verification/exp4/README.md is written in an unsupported language. File is not indexed.
view on githubraw file Latest commit 7fe51dd4 on 2025-03-14 19:25:07 UTC
7fe51dd492 Jean*0001 Example: "Flow over a bump with Open Boundaries and passive tracers"
0002 ====================================================================
0003
0004 This experiment is a 400 x 210 x 4.5 km channel (80 x 42 x 8 grid-points)
0005 on an f-plane with a tall seamount in the middle.
0006 It is intended to illustrate the interaction of a zonal flow over and around
0007 a steep bump as well as the use of Open Boundary (`pkg/obcs`) and Floats
0008 packages (`pkg/flt`).
0009
0010 ### Overview:
0011 This experiment contains 4 set-ups (with corresponding `input[.*]/` dir) that
0012 can be run with the same executable (built from `build/` dir using customized
0013 code from `code/`);
0014 binary input files (all `real*8`) have been generated using matlab script
0015 `gendata.m` from `input` dir.
0016 All four set-ups use a simple EOS ( $\rho' = -\rho_0 ~ \alpha_T ~ \theta'$ )
0017 and treat salt as a passive tracer ;
0018
0019 The **primary** test, using input files from `input/` dir, use four open
0020 boundaries with simple specifications (`useOBCSprescribe`) from open boundary
0021 parameter file `data.obcs`.
0022 This is a non-hydrostatic set-up using the flux-form momentum equations.
0023
0024 Different kinds of open boundary values are used:
0025 zonal (x-)velocity U is prescribed at all open boundaries with values that are
0026 read from data files (specified in data.obcs);
0027 meridional (y-)velocity V is set to zero on all boundaries, and temperature to
0028 `tRef(z)`, both `in obcs_calc.F`, this is the default behavior;
0029 at the western boundary, salinity values are used for salinity and one passive
0030 tracer in the same way.
0031 Salinity is set to sLev at all other boundaries, while a (nearly) homogeneous
0032 Neumann condition is applied to the passive tracer (the latter is the default
0033 in `obcs_calc.F`), with a relaxation (using pkg rbcs) in the Eastern part of
0034 the channel.
0035
0036 The **secondary** test, using `input.nlfs/` dir, is similar to the primary test except
0037 it is hydrostatic with the vector-invariant momentum formulation and using the
0038 `z*` coordinate. A time-varying small imbalance between the Western boundary
0039 inflow and Eastern boundary outflow generates sea-level fluctuations.
0040
0041 The **secondary** test, using `input.stevens/` dir, is very similar to the primary test
0042 except it is hydrostatic with some bottom friction and use Stevens Open-Boundary
0043 Condition formulation at the Eastern and Western edges.
0044
0045 The **secondary** test, using `input.with_flt/` dir, is intended to illustrate the use
0046 of `pkg/flt` in a simple configuration (hydrostatic, no Open-Boundary but force
0047 with a zonal wind-stress instead).
0048 ***Note:*** this set-up has been moved (in PR #830) from `verification/flt_example/`
0049
0050 ### Instructions:
0051 default paths are (as used in `testreport`):
0052
0053 ```
0054 set build_dir = ${cwd}/build
0055 set run_dir = ${cwd}/run
0056 set MITGCM = ${cwd}/../../
0057 ```
0058
0059 Configure and compile the code:
0060
0061 ```
0062 cd ${build_dir}
0063 ${MITGCM}/tools/genmake2 -mods ../code [-of my_platform_optionFile]
0064 [make Clean]
0065 make depend
0066 make
0067 cd ..
0068 ```
0069
0070 To run primary test:
0071
0072 ```
0073 cd ${run_dir}
0074 ln -s ../input/* .
0075 ${build_dir}/mitgcmuv > output.txt
0076 cd ..
0077 ```
0078
0079 There is comparison output in the directory:
0080 results ( absolute path: `${run_dir}/../results` )
0081
0082 To run any of secondary `$st` test (`$st` in: `nlfs`, `stevens`, `with_flt`):
0083
0084 ```
0085 cd ${run_dir}
0086 rm *
0087 ln -s ../input.$st/* .
0088 ln -s ../input/* .
0089 ${build_dir}/mitgcmuv > output.txt
0090 cd ..
0091 ```
0092
0093 There is comparison output in the directory:
0094
0095 `results/output.$st.txt`
0096
0097 ### Notes:
