Run the coupled case with WaveWatch III
Case initialization
To install the coupled case with WaveWatch III, enter the L4C1 case folder
cd $SKRIPS_DIR/coupler/L4.C1.coupled_RS2012_ring/
./install.sh
Initialize the test case
cd runCase.init
./Allrun
Run the coupled case
Enter the runCase folder and run the test case:
[ruisun@acc00]~/.../coupler/L4.C1.coupled_RS2012_ring/runCase.init$ cd ../runCase/
[ruisun@acc00]~/.../coupler/L4.C1.coupled_RS2012_ring/runCase$ ./Allrun
Similar with the L3C1 case, the namelist.rc file controls the coupled run:
DebugLevel: 0
## mode 1 = sequential mode
## mode 2 = concurrent mode
coupleMode: 1
cpuOCN: 16
cpuATM: 16
cpuWAV: 16
## Set NX and NY for WaveWatch III in ESMF grid
## NX * NY = cpuWAV
waveNPX: 4
waveNPY: 4
StartYear: 2012
StartMonth: 06
StartDay: 01
StartHour: 00
StartMinute: 00
StartSecond: 00
StopYear: 2012
StopMonth: 06
StopDay: 01
StopHour: 03
StopMinute: 00
StopSecond: 00
## EsmStepSeconds is the coupling time step
EsmStepSeconds: 600
## ATMStepSeconds is the time step for WRF
ATMStepSeconds: 60
## OCNStepSeconds is the time step for MITgcm
OCNStepSeconds: 600
## The wave model is not activated
WAVStepSeconds: 600
In namelist.input file, I added the following lines to control the input and output from ESMF:
## auxinput5_interval_s sets the coupling interval for ESMF inputs
30 auxinput5_inname = 'wrfin_esmf',
31 auxinput5_interval_s = 60,
32 auxinput5_end_d = 60,
33 io_form_auxinput5 = 7,
## auxhist5_interval_s sets the coupling interval for ESMF outputs
34 auxhist5_outname = 'wrfout_esmf',
35 auxhist5_interval_s = 60,
36 auxhist5_end_d = 60,
37 io_form_auxhist5 = 7,
In namelist.input file, I added the following lines to control the coupling schemes for WRF:
## Option 6: Chanock;
## Option 7: Taylor and Yelland;
## Option 8: OOST;
86 isftcflx = 7,
## Option 0: use COARE 3.0 scheme;
## Option 1: use COARE 3.5 scheme;
87 use_coare35 = 0,
In data file, I added the following lines to control coupling scheme for MITgcm:
44 # Options when coupled with WW3
## Option 0: use wind stress from MITgcm;
## Option 1: use wind stress from WW3;
41 stressFromWave = 0,
## Option 0: do not activate Langmuir turbulence in KPP;
## Option 1: use LF16-MA;
## Option 2: use LF16-EN;
## Option 3: use LF17;
## Option 5: use LF17, but with parameterized wave;
42 langmuirScheme = 0,
## Option 0: do not activate the Stokes-Coriolis or Stokes advection;
## Option 1: use Stokes-Coriolis and Stokes advection;
## To activate Stokes advection, MITgcm MUST have multiDimAdvection = .TRUE.
## To activate Stokes-Coriolis, MITgcm MUST have vectorInvariantMomentum = .FALSE.
## stokesProfile = 0: no stokes profile (turn off stokesCoriolis and stokesAdvection)
## stokesProfile = 1: use Eq. (7) in Breivik 2014 paper (https://doi.org/10.1175/JPO-D-14-0020.1)
## stokesProfile = 2: use Eq. (16) in Breivik 2014 paper
43 stokesProfile = 0,
44 stokesCoriolis = 0,
45 stokesAdvection = 0,