kshedstrom/roms

Regional Ocean Modeling System (with ice)

roms's Issues

Bug in calculation of grid spacing in roms_to_pop.py

I just discovered a bug in the python program roms_to_pop.py that is used to create a model grid file for CICE based on a ROMS grid. Lines 50-53 read:

htn = 0.5*(dy[1:-1,1:-1] + dy[2:,1:-1])
hus = 0.5*(dx[1:-1,1:-1] + dx[1:-1,2:])
hte = 0.5*(dy[1:-1,1:-1] + dy[1:-1,2:])
huw = 0.5*(dx[1:-1,1:-1] + dx[2:,1:-1])

Based on the CICE documentation, the variable htn is the "length of northern edge (Δx) of T-cell", however, in the code above it is set to the Δy of the ROMS grid. This should not matter too much if the grid cells are approximately square in a kilometer sense, but will make a difference if the grid cells are more rectangular.

ICE_DIAGS FLAG Repeats

I believe this is some sort of Copy/Paste mistake. There is are duplicates of ICE_DIAGS flag in the bulk_flux.F which results in compile failure due to repeat definitions.

roms/ROMS/Nonlinear/bulk_flux.F

Line 214 in 0baf667

# ifdef ICE_DIAGS

# ifdef ICE_DIAGS
     &                     FORCES(ng) % LW_down,                        &
     &                     FORCES(ng) % SW_down,                        &
     &                     FORCES(ng) % lat_ice,                        &
     &                     FORCES(ng) % sens_ice,                       &
     &                     FORCES(ng) % LW_up_ice,                      &
     &                     FORCES(ng) % SW_up_ice,                      &
     &                     FORCES(ng) % saltflux_ocean,                 &
# endif
# ifdef ICE_DIAGS
     &                     FORCES(ng) % LW_down,                        &
     &                     FORCES(ng) % SW_down,                        &
     &                     FORCES(ng) % lat_ice,                        &
     &                     FORCES(ng) % sens_ice,                       &
     &                     FORCES(ng) % LW_up_ice,                      &
     &                     FORCES(ng) % SW_up_ice,                      &
     &                     FORCES(ng) % saltflux_ocean,                 &

Potential tides - not reading from forcing file

Hi Kate,

I am not sure whether this is an issue or just a change in how things are done between 3.7 and 3.6.

I am currently using stefanie mack's 3.7 branch (with ice shelves in it) but it was based off of your branch. And it looks like it has the potential tide code in it, but I can't get it to read the tide forcing file properly.

I was wondering if you have run a model using potential tides using your version of the roms3.7 model?

The tide forcing file I use works in 3.6. and it looks like it can't be read in the 3.7 version.
It fails when it should be reading in the potential tide amp: From the *.out file

    GET_2DFLD   - maximum tidal current, ellipse major axis
                   (Grid = 01, Min =  0.00000000E+00 Max =  2.01598481E+01)
   GET_2DFLD   - minimum tidal current, ellipse minor axis
                   (Grid = 01, Min = -2.97327515E+00 Max =  2.01632515E+00)


 INQUIRY     - unable to find requested variable: \00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00

In the 3.6 code, it reads it in and it all works well.

GET_2DFLD   - maximum tidal current, ellipse major axis
               (Min =  0.00000000E+00 Max =  2.02808533E+01)
GET_2DFLD   - minimum tidal current, ellipse minor axis
               (Min = -2.97327515E+00 Max =  2.16472321E+00)
GET_2DFLD   - tidal potential elevation amplitude
               (Min =  0.00000000E+00 Max =  1.23815612E-01)
GET_2DFLD   - tidal potential elevation phase angle
               (Min =  0.00000000E+00 Max =  6.28318517E+00)

I will keep looking at this more see if I can tell what is going on in the code, but wanted to know if you had any ideas, and to see whether you have successfully used potential tide with roms3.7.

bug in ccsm_flux

Hi @kshedstrom !

I was working on yet another implementation of LY2004 and looking at the ROMS code, something
stroke me as suspicious. The equation for the neutral latent coef in LY2004 eq 6b is

1000 Ce = 34.6 * sqrt(Cd)

but in the code, it became:

Ce = 34.6 /1000

in 3 places:

roms/ROMS/Nonlinear/ccsm_flux.F

Line 860 in cc7adbf

ren = 0.0346_r8

roms/ROMS/Nonlinear/ccsm_flux.F

Line 884 in cc7adbf

ren = 0.0346_r8

roms/ROMS/Nonlinear/ccsm_flux.F

Line 918 in cc7adbf

ren = 0.0346_r8

Am I missing something or did we carry a bug all those years?

albedo.F90

Hi Kate. Seems variable for albedo over ocean 'alb_w' is not defined unless you are running ROMS with the ice model. In albedo.F90 I had to move

real(r8), parameter :: alb_w=0.06_r8

outside of the ifdefs:
ifdef ICE_BULK_FLUXES
ifdef ALBEDO_CSIM