Particles outside of the grid nest warning

Guardiola mélanie

unread,

Sep 17, 2019, 5:00:54 AM9/17/19

to connectivity-modeling-system-club

Dear all,

I am using CMS to compute trajectories of particles inside the Mediterranean Sea using oceanographic model (Mercator PSY2V4R4). This is a non-regular C-grid.

According to the cms user guide, lon and lat have to be in 1D for the cms code. So I interpolated U, V and W at the center of the grid cell in order to get a A-grid and then I interpolated it on a regular one (please find attached an example file).

From Matlab, here is an ncdisp of my file :

Format:
           classic
Global Attributes:
           grid_type = 'curvilinear'
Dimensions:
           x    = 617
           y    = 473
           z    = 50
           time = 1     (UNLIMITED)
Variables:
    Longitude
           Size:       617x1
           Dimensions: x
           Datatype:   single
           Attributes:
                       units = 'degree_north'
    Latitude
           Size:       473x1
           Dimensions: y
           Datatype:   single
           Attributes:
                       units = 'degree_east'
    Depth
           Size:       50x1
           Dimensions: z
           Datatype:   single
           Attributes:
                       units = 'meter'
    zu
           Size:       617x473x50x1
           Dimensions: x,y,z,time
           Datatype:   single
           Attributes:
                       long_name = 'Zonal current'
                       units      = 'm/s'
                       _FillValue = 1.267650600228229e+30
    zv
           Size:       617x473x50x1
           Dimensions: x,y,z,time
           Datatype:   single
           Attributes:
                       long_name = 'Meridional current'
                       units      = 'm/s'
                       _FillValue = 1.267650600228229e+30
    zw
           Size:       617x473x50x1
           Dimensions: x,y,z,time
           Datatype:   single
           Attributes:
                       long_name = 'Vertical current'
                       units      = 'm/s'
                       _FillValue = 1.267650600228229e+30

But I am experiencing an issue with the particles released during experiment. I made sure their starting position is located inside the input grid

1) but with this releaseFile :

1   -11 36.000 0.6 1 2007 1 1 0
1    -8 36.000 0.6 1 2007 1 1 0
1    -3 36.000 0.6 1 2007 1 1 0
1   10 38.000 0.6 1 2007 1 1 0

when I run the CMS code (./cms …) I get the output :

==========================================
File name : traj_file_1.nc
Total number of release events     : 4 (numbers 1 to 4)
Total number of particles released : 4
Total number of time steps         : 13
Total number of time steps in output file: 13
==========================================
WARNING: Release line           1 is outside of the nest grid
WARNING: Release line           2 is outside of the nest grid
WARNING: Release line           3 is outside of the nest grid

So it seems negatives longitudes doesn’t work

2) but if I try this releaseFile for example :

1 -11 36.000 0.6 1 2007 1 1 0

1 -8 36.000 0.6 1 2007 1 1 0

1 3 38.000 0.6 1 2007 1 1 0

1 10 38.000 0.6 1 2007 1 1 0

I get :

==========================================

File name : traj_file_1.nc

Total number of release events : 4 (numbers 1 to 4)

Total number of particles released : 4

Total number of time steps : 13

Total number of time steps in output file: 13

==========================================

WARNING: Release line 1 is outside of the nest grid

WARNING: Release line 2 is outside of the nest grid

WARNING: Release line 4 is outside of the nest grid

So the line 4 which was previously « ok » isn’t anymore even though I haven’t changed this line in the second experiment.

In any of these two cases the the trajectory for the particles isn't computed because I get for lon and lat:

1) lon

349 352 -3 10

NaN NaN NaN 10

NaN NaN NaN NaN

lat

36 36 36 10

NaN NaN NaN 10

NaN NaN NaN NaN

2) lon

349 352 3 10

NaN NaN 3 NaN

NaN NaN NaN NaN

lat

36 36 38 10

NaN NaN 38 NaN

NaN NaN NaN NaN

I guess the issue is linked to the structure of my input but I haven’t figured out what it is about yet. Maybe some of you have experienced similar issue ?

Thanks a lot for your help

Regards,

Mélanie

ibm.list

nest_1.nml

runconf.list

Guardiola mélanie

unread,

Oct 1, 2019, 3:52:20 AM10/1/19

to connectivity-modeling-system-club

Hi everyone,

I found out what caused the issue. It is due to the number of nests files in the 'nests' directory. I just wanted to try cms whith only 1 or 2 nests input files to make sure it works with before processing the whole period (which takes a bunch of time).

But I noticed the error disappears if I use at least 3 nests input files.

Hope it can help.

Mélanie

Robin F

unread,

Oct 1, 2019, 9:02:49 AM10/1/19

to Guardiola mélanie, connectivity-modeling-system-club

Hi Mélanie,

Thanks for the update. Indeed, you need to provide a few time steps before the release of your particles so that the CMS can interpolate the currents to the particles’ locations.

Best,

--

Robin Faillettaz, Ph.D
——————————————————————

Postdoctoral Research Fellow
RECOVER2 project
University of Miami
Rosenstiel School of Marine & Atmospheric Science
4600 Rickenbacker Causeway, Miami, FL 33149-1098, USA
Tel: (US) +1-702-910-0380; (FR) +33-6-84-91-43-33
Robin.Fa...@gmail.com
Robin.Fa...@rsmas.miami.edu

Reply all

Reply to author

Forward