#2561 closed help (answered)

TRIP 1D, 2D

Reported by: pmcguire Owned by: pmcguire
Component: JULES Keywords: TRIP, 1D Land Only Compression
Cc: Platform:
UM Version:

Description

Hi Patrick and Pier Luigi,

The trip at 1deg simulation is running. However, I have a doubt
regarding some parameters:

The documentation (https://goo.gl/UhuZNW) says that nx_grid, ny_grid,
reg_dlat, reg_dlon, reg_lat1, reglon1 should be defined just if
grid_is_1d=.true. .

As we have grid_is_1d=.false., I tried to run without these parameters
but the simulation failed. It just worked when all of them are defined.
So, my final configuration is:

[namelist:jules_input_grid]
grid_is_1d=.false.

[namelist:jules_rivers_props]
file='/home/users/ovmuller/tripdata/trip_dir_seq.1deg.nc'
nvars=2
use_file=.true.,.true.
var='direction','sequence'
var_name='direction','sequence'
nx=360
ny=180
x_dim_name='lon'
y_dim_name='lat'
rivers_regrid=.true.
nx_grid=720
ny_grid=360
reg_dlat=0.5
reg_dlon=0.5
reg_lat1=-89.75
reg_lon1=-179.75

Can you explain it? I mean, this configuration is not coincident with
the documentation. I guess that maybe the:

[namelist:jules_model_grid]
force_1d_grid=.false.
land_only=.true.

is forcing grid_is_1d to be true.

Has it sense?
Omar Mueller

Change History (2)

comment:1 Changed 14 months ago by pmcguire

Hi Omar

You are looking at Jules 4.2 documentation. The Jules 4.9 documentation says that they need to be defined in that case or for a parallel standalone run. Since we are doing parallel standalone runs, they need to be defined.
I think you must run with grid_is_1d=false, since for example, the other ancillary files are 2D. The output grid I have been normally using is 1D, since it runs faster with smaller output files. You can try 2D output if you'd like. It's not hard to set it up for that.

Patrick

comment:2 Changed 14 months ago by pmcguire

  • Resolution set to answered
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.