GET_2DFLD_NF90 - error while reading variable: Uwind at TIME index = 56

[iriskeizer]

Hello everyone,

I'm a beginner at working with ROMS. Currently, I'm trying to run an application for the NWES very similar to the one used in this paper: https://doi.org/10.1029/2020JC016325
However, with the newest version of the model.

When I try to run my application on the Snellius supercomputer, it crashes at TIME index 56 while reading variable Uwind.
I run the model from 1993-01-01. The Uwind, Vwind, Tair, Pair and Qair are forced 6-hourly at 00:00, 06:00, 12:00 and 18:00.
Total precipitation is forced 12-hourly at 06:00 and 18:00 and solar radiation is forced daily at 12:00.

What is special about TIME index 56 is that it is 1993-01-15, 00:00. This means that from the boundary, climatology and river forcing files no longer the value from 1992-12-15 00:00:00.00 should be used but for the first time the value from '1993-01-15T00:00:00.000000000'. However, when I look in the log file I see this:

Part of log file:
GET_STATE_NF90 - NLM: state initial conditions, 1993-01-15 00:00:00.00
(Grid 01, t = 16451.0000, File: NorthSea8_rst_Y2019_14_01_2018_NoneVar.nc, Rec=0001, Index=1)
- free-surface
(Min = 0.00000000E+00 Max = 0.00000000E+00)
- vertically integrated u-momentum component
(Min = 0.00000000E+00 Max = 0.00000000E+00)
- vertically integrated v-momentum component
(Min = 0.00000000E+00 Max = 0.00000000E+00)
- u-momentum component
(Min = 0.00000000E+00 Max = 0.00000000E+00)
- v-momentum component
(Min = 0.00000000E+00 Max = 0.00000000E+00)
- potential temperature
(Min = 0.00000000E+00 Max = 0.00000000E+00)
- salinity
(Min = 0.00000000E+00 Max = 0.00000000E+00)
- vertical viscosity coefficient
(Min = 0.00000000E+00 Max = 0.00000000E+00)
- temperature vertical diffusion coefficient
(Min = 0.00000000E+00 Max = 0.00000000E+00)
- salinity vertical diffusion coefficient
(Min = 0.00000000E+00 Max = 0.00000000E+00)
GET_NGFLD_NF90 - river runoff XI-positions at RHO-points
(Grid = 01, Min = 8.50000000E+01 Max = 2.40000000E+02)
GET_NGFLD_NF90 - river runoff ETA-positions at RHO-points
(Grid = 01, Min = 6.00000000E+00 Max = 1.85000000E+02)
GET_NGFLD_NF90 - river runoff direction
(Grid = 01, Min = 0.00000000E+00 Max = 0.00000000E+00)
GET_NGFLD_NF90 - river runoff mass transport vertical profile
(Grid = 01, Min = 0.00000000E+00 Max = 1.00000000E+00)
GET_NGFLD_NF90 - river runoff potential temperature, 1992-12-15 00:00:00.00
(Grid= 01, Rec=540, Index=2, File: Rivers_NorthSea8_smooth013_Dai2014_AnnualCycle_1993-2019.nc)
(Tmin= 14.0000 Tmax= 26281.0000) t = 16420.0000
(Min = 0.00000000E+00 Max = 1.39565896E+01)
GET_NGFLD_NF90 - river runoff salinity, 1992-12-15 00:00:00.00
(Grid= 01, Rec=540, Index=2, File: Rivers_NorthSea8_smooth013_Dai2014_AnnualCycle_1993-2019.nc)
(Tmin= 14.0000 Tmax= 26281.0000) t = 16420.0000
(Min = 0.00000000E+00 Max = 0.00000000E+00)


For some reason for the river, still, 1992-12-15 00:00:00.00 is used. How might this be caused? Is this correct? How can I fix this?
Is this also the reason for the crash or is there a different reason?



Statement of problem in log:
REGRID - input gridded data does not contain model grid:

Gridded: LonMin = -24.0000 LonMax = 14.0000
LatMin = 32.0000 LatMax = 66.0000
Model: LonMin = 0.0000 LonMax = 0.0000
LatMin = 0.0000 LatMax = 0.0000
Found Error: 04 Line: 354 Source: ROMS/Utility/get_2dfld.F, get_2dfld_nf90

GET_2DFLD_NF90 - error while reading variable: Uwind at TIME index = 56
Found Error: 04 Line: 112 Source: ROMS/Utility/get_2dfld.F
Found Error: 04 Line: 169 Source: ROMS/Nonlinear/get_data.F
Found Error: 04 Line: 670 Source: ROMS/Nonlinear/initial.F
Found Error: 04 Line: 197 Source: ROMS/Drivers/nl_roms.h, ROMS_initialize




End of log:
<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
Found Error: 04 Line: 85 Source: ROMS/Utility/close_io.F, close_file_nf90
Found Error: 04 Line: 125 Source: ROMS/Utility/close_io.F, close_inp
Found Error: 04 Line: 85 Source: ROMS/Utility/close_io.F, close_file_nf90
Found Error: 04 Line: 85 Source: ROMS/Utility/close_io.F, close_file_nf90
Found Error: 04 Line: 85 Source: ROMS/Utility/close_io.F, close_file_nf90

ROMS/TOMS - Output NetCDF summary for Grid 01:
Found Error: 04 Line: 401 Source: ROMS/Utility/close_io.F, close_out

ROMS/TOMS - I/O error ............... exit_flag: 4

[wilkin]

The error says ROMS is failing while attempting to regrid the Uwind data to the model grid:

Statement of problem in log:
REGRID - input gridded data does not contain model grid:

Gridded: LonMin = -24.0000 LonMax = 14.0000
LatMin = 32.0000 LatMax = 66.0000
Model: LonMin = 0.0000 LonMax = 0.0000
LatMin = 0.0000 LatMax = 0.0000
Found Error: 04 Line: 354 Source: ROMS/Utility/get_2dfld.F, get_2dfld_nf90

GET_2DFLD_NF90 - error while reading variable: Uwind at TIME index = 56

The clue here is the range of model lon/lat in the grid file, which appears to be zero to zero.

Look at the log file report for the grid input starting right after the line ...

INITIAL: Configuring and initializing forward nonlinear model .

Aside from bathymetry which ranges from 0 to 5.417E+03, all other grid variables are zero:

GET_GRID_NF90 - Coriolis parameter at RHO-points: f
(Grid = 01, File: NorthSea8_smooth013_sponge_nudg.nc)
(Min = 0.00000000E+00 Max = 0.00000000E+00)
GET_GRID_NF90 - reciprocal XI-grid spacing: pm
(Grid = 01, File: NorthSea8_smooth013_sponge_nudg.nc)
(Min = 0.00000000E+00 Max = 0.00000000E+00)
...
GET_GRID_NF90 - x-location of RHO-points: x-rho
(Grid = 01, File: NorthSea8_smooth013_sponge_nudg.nc)
(Min = 0.00000000E+00 Max = 0.00000000E+00)
GET_GRID_NF90 - y-location of RHO-points: y_rho
(Grid = 01, File: NorthSea8_smooth013_sponge_nudg.nc)
(Min = 0.00000000E+00 Max = 0.00000000E+00)

pm = pn = 0 means your grid spacings delta-x,delta-y are infinite.

For your application, ROMS should be reading lon/lat coordinates, not x_rho,y_rho. You have CURVGRID defined, so it's not that. Is variable spherical set in the grid file?

[wilkin]

Also check your rivers options. The log says ...

Rivers_NorthSea8_smooth013_Dai2014_AnnualCycle_1993-2019.nc
LuvSrc = F u-face = 63
LuvSrc = F v-face = 0
LwSrc = F w-face = 0

You have FALSE for all, which I expect renders the rivers inactive.

There were recent enhancements to the LuvSrc/LwSrc options to allow a mix of both, not just all rivers with one method. See this release note:
https://www.myroms.org/projects/src/ticket/905

[iriskeizer]

Thanks, indeed, I had to change river_direction flags from 0 to 2 for all rivers.

[iriskeizer]

The variables where zero because in the grid file I was using those variables were zero.

Thanks, for helping out!