Open kyle-hinson opened 2 months ago
jcwarner-usgs
commented
2 months ago you need to make changes in Master/mct_roms_wrf.h and in WRF/frame/atm_coupler.F Look at SST as an example. You might need to add something into WRF/Registry. not sure. WRF will continue to build after an error. ROMS will stop if it gets a error. because WRF uses make -? (i forget the command, -k or something). Is grid%ucurrent already a WRF variable? i think you want to hook into grid%uoce and grid%voce in wrf.
kyle-hinson
commented
2 months ago Thanks for the feedback! I had already made changes to Master/mct_roms_wrf.h and WRF/frame/atm_coupler.F, but it looks like once I changed the variable names from 'UCURRENT' and 'VCURRENT' to 'UOCE' and 'VOCE' that fixed many of the problems and I was able to compile successfully.
I think that the main issue that I'm having now is related to my modifications to WRF/phys/module_sf_mynn.F so that these currents actually interact with surface-level winds. I'm attempting to add the surface currents to variables like U10 and V10 (example modifications below).
!-----------------------------------------------------
!COMPUTE 10 M WNDS
!-----------------------------------------------------
! If the lowest model level is close to 10-m, use it
! instead of the flux-based diagnostic formula.
if (ZA(i) .le. 7.0) then
! high vertical resolution
if(ZA2(i) .gt. 7.0 .and. ZA2(i) .lt. 13.0) then
!use 2nd model level
U10(I)=U1D2(I)+UOCE(I)
V10(I)=V1D2(I)+VOCE(I)
else
U10(I)=(U1D(I)+UOCE(I))*log(10./ZNTstoch(I))/log(ZA(I)/ZNTstoch(I))
V10(I)=(V1D(I)+VOCE(I))*log(10./ZNTstoch(I))/log(ZA(I)/ZNTstoch(I))
endif
elseif(ZA(i) .gt. 7.0 .and. ZA(i) .lt. 13.0) then
!moderate vertical resolution
!U10(I)=U1D(I)*PSIX10/PSIX
!V10(I)=V1D(I)*PSIX10/PSIX
!use neutral-log:
U10(I)=(U1D(I)+UOCE(I))*log(10./ZNTstoch(I))/log(ZA(I)/ZNTstoch(I))
V10(I)=(V1D(I)+VOCE(I))*log(10./ZNTstoch(I))/log(ZA(I)/ZNTstoch(I))
else
! very coarse vertical resolution
U10(I)=(U1D(I)+UOCE(I))*PSIX10/PSIX
V10(I)=(V1D(I)+VOCE(I))*PSIX10/PSIX
endifHowever, the error that I'm getting seems to be related to an inability to define this variable in the SFCLAY1D_mynn submodules. Is this what has to be modified in the WRF registry (maybe 'WRF/Registry/Registry'?) file that you mentioned?
module_sf_mynn.f90(738): error #6420: This array name is invalid in this context. [UOCE]
REAL, DIMENSION( ims:ime ), INTENT(IN) :: UOCE, VOCE
-------------------------------------------------^
module_sf_mynn.f90(738): error #6420: This array name is invalid in this context. [VOCE]
REAL, DIMENSION( ims:ime ), INTENT(IN) :: UOCE, VOCE
-------------------------------------------------------^
module_sf_mynn.f90(612): error #6784: The number of actual arguments cannot be greater than the number of dummy arguments. [SFCLAY1D_MYNN]
CALL SFCLAY1D_mynn( &
-------------^
module_sf_mynn.f90(631): error #6633: The type of the actual argument differs from the type of the dummy argument. [UOCE]
UOCE(ims,j), VOCE(ims,j), &
----------------^
module_sf_mynn.f90(631): error #6633: The type of the actual argument differs from the type of the dummy argument. [VOCE]
UOCE(ims,j), VOCE(ims,j), &
-----------------------------^
module_sf_mynn.f90(635): error #6633: The type of the actual argument differs from the type of the dummy argument. [ISFTCFLX]
,isftcflx,iz0tlnd, &
-----------------^
module_sf_mynn.f90(635): error #6633: The type of the actual argument differs from the type of the dummy argument. [IZ0TLND]
,isftcflx,iz0tlnd, &
--------------------------^
module_sf_mynn.f90(635): error #6780: A dummy argument with the INTENT(IN) attribute shall not be defined nor become undefined. [ISFTCFLX]
,isftcflx,iz0tlnd, &
-----------------^
module_sf_mynn.f90(635): error #6780: A dummy argument with the INTENT(IN) attribute shall not be defined nor become undefined. [IZ0TLND]
,isftcflx,iz0tlnd, &
--------------------------^
module_sf_mynn.f90(612): error #8284: If the actual argument is scalar, the dummy argument shall be scalar unless the actual argument is of type character or is an element of an array that is not assumed shape, pointer, or polymorphic. [USTM]
CALL SFCLAY1D_mynn( &
-------------^
module_sf_mynn.f90(612): error #8284: If the actual argument is scalar, the dummy argument shall be scalar unless the actual argument is of type character or is an element of an array that is not assumed shape, pointer, or polymorphic. [CK]
CALL SFCLAY1D_mynn( &
-------------^
compilation aborted for module_sf_mynn.f90 (code 1)you might need to add the UOCS and VOCE to the routine that calls sfclay. you will need to go backwards in the code.
kyle-hinson
commented
2 months ago Thanks for the suggestion. This is what I had previously added into the sfclay routine, trying to follow other variables like HWAVE, LWAVEP, and PWAVE.
!-------------------------------------------------------------------
SUBROUTINE SFCLAY_mynn( &
U3D,V3D,T3D,QV3D,P3D,dz8w, &
CP,G,ROVCP,R,XLV,PSFCPA,CHS,CHS2,CQS2,CPM, &
ZNT,UST,PBLH,MAVAIL,ZOL,MOL,REGIME,PSIM,PSIH, &
XLAND,HFX,QFX,LH,TSK,FLHC,FLQC,QGH,QSFC,RMOL, &
U10,V10,TH2,T2,Q2,SNOWH, &
GZ1OZ0,WSPD,BR,ISFFLX,DX, &
SVP1,SVP2,SVP3,SVPT0,EP1,EP2, &
KARMAN,itimestep,ch,th3d,pi3d,qc3d,rho3d,qcg, &
spp_pbl,pattern_spp_pbl, &
#if defined SWAN_COUPLING || defined WW3_COUPLING
HWAVE, LWAVEP, PWAVE, &
#endif
!======ADDED SURFACE CURRENTS======================
#if defined ROMS_COUPLING
UOCE, VOCE, &
#endif
!==================================================
ids,ide, jds,jde, kds,kde, &
ims,ime, jms,jme, kms,kme, &
its,ite, jts,jte, kts,kte, &
ustm,ck,cka,cd,cda,isftcflx,iz0tlnd )
!-------------------------------------------------------------------I've also already added these variables in the additional SFCLAY_1D_MYNN subroutine. Are you saying that this function is defined somewhere else in WRF as well? Or am I misunderstanding where else these variables need to be declared?
kyle-hinson
commented
1 month ago Was able to track down some of my issues that I previously ran into (missing comma threw a syntax error), but it looks as if the new error is pointing me to WRF/phys/module_surface_driver.F as the new source of my issues. UOCE and VOCE are both already defined in this file, and I'm not entirely sure what to make of the error below.
module_surface_driver.f90(5146): error #6404: This name does not have a type, and must have an explicit type. [UOCE]
UOCE, VOCE, &
----------^
module_surface_driver.f90(5146): error #6404: This name does not have a type, and must have an explicit type. [VOCE]
UOCE, VOCE, &
----------------^
compilation aborted for module_surface_driver.f90 (code 1)Do I need to create a new set of variables like SST_INPUT for UOCE and VOCE while changing the definitions from
REAL, DIMENSION( ims:ime, jms:jme ), INTENT(OUT):: UOCE
REAL, DIMENSION( ims:ime, jms:jme ), INTENT(OUT):: VOCEto something more like what I have below so that they can have similar cpl_rcv functions as SST later in module_surface_driver.F? I tried compiling with this code setup without any luck so far.
REAL, DIMENSION( ims:ime, jms:jme ), INTENT(INOUT):: UOCE
REAL, DIMENSION( ims:ime, jms:jme ), INTENT(INOUT):: VOCE
REAL, DIMENSION( ims:ime, jms:jme ), INTENT(IN ):: UOCE_INPUT
REAL, DIMENSION( ims:ime, jms:jme ), INTENT(IN ):: VOCE_INPUTi dont think you should make new vars. there is just something off with the current uoce and voce. can you post your moduels_surface_driver.F
kyle-hinson
commented
1 month ago Sure thing, attached here. Have also included my module_sf_mynn.F if that's useful.
phys_modules.zip
Hi all. I'm trying to compile a ROMS-WRF setup (no waves) in COAWST that includes the effects of ROMS surface currents on WRF surface winds. I've successfully compiled a 'base case' ROMS-WRF setup, and have tried to modify the code to exchange information about surface currents in Master/roms_wrf_mct.h, WRF/frame/atm_coupler.F, and WRF/phys/module_sf_mynn.F (attached files include edits following work done by Shi and Bourassa, 2019).
coawst_modifications.zip
I'm not sure exactly what additional modifications to the code I might be missing, below are a couple snippets of output from my attempt to compile. Here it seems to run into some major issues, but continues compiling:
And below is where it finally finishes attempting to compile and fails.
I looked around a bit at Hernan's recent test case to see what other modifications were made to the code but wasn't sure if this has been fully integrated into COAWST already. Any pointers as to where else I can define the new 'ucurrent' and 'vcurrent' variables to make sure that they're exchanged with the atmospheric coupler or MCT exchange would be much appreciated, thanks for any help.