mitgcm / gcmfaces

gcmfaces is a Matlab / Octave toolbox that handles gridded earth variables in generic fashion. Read more at:

MITgcm

MIT General Circulation Model master code and documentation. The main MITgcm webpage can be found here.

Documentation

Access the latest documentation here

gcmfaces's Issues

can't use grid_load in matlab

I have put the nctiles_grid and gcmfaces in a same directory according to the v4r4 user guider. But when I run the code with ECCOv4r4 like this:
p = genpath('gcmfaces/'); addpath(p);
p = genpath('MITprof/'); addpath(p);
grid_load
it will show that 'Unable to open file nctiles_grid\GRID.0001.nc'.
The code is OK when I use ECCO v4r3. So how can I solve it? Thank you.

Question about the use of read_nctiles

I have copy the update commit 'read_nctiles', but when I read the variables to calculate the volume budget like this:
oceFWflx = read_nctiles( [dir1,'oceFWflx/','oceFWflx'],'oceFWflx',ii )
It will show that '/ECCOv4r4/nctiles_monthly/oceFWflx/oceFWflx.0001.nc not found in the current directory'.
Actually, I noticed that the file named by oceFWflx.0001.nc didn't exist in the ECCOv4r4, in which the files are stored in sub-directories of different year and named by the months, rather than divided into 13 tiles like v4r3.
So how can I solve it and read the variables successfully to calculate the volume budget with v4r4?

change the default tile size in prep_flt_init.m

Hi, the code "prep_flt_init.m" set the tile size as 3030. But my llc540 simulation use the tile of 6060. So I should replace all "30" with "60" in the "prep_flt_init.m" to repare my initial flt files, am I right? Many thanks!

need to support nctiles variant used in ECCO v4 r4

For ECCO v4 r4, a slightly different version of nctiles was used, which is not yet supported by gcmfaces read_nctiles.m.

New format looks like this:

	float UVELMASS(time, k, tile, j, i_g) ;
		UVELMASS:units = "m/s" ;
		UVELMASS:long_name = "X-Comp of Geometry-Weighted Velocity (m/s)" ;
		UVELMASS:coordinates = "timestep Z" ;

@owang1 , do you think that read_nctiles.m could easily be modified to support both the newer and older nctiles files?

Interpolating llc output to a regular lat-lon grid

Is there a routine for interpolating llc output (both 2d and 3d) to a regular lat-lon grid? Thanks!

Flux conserving, bin average of 2-D field

Hi Gaël,

Does gcmfaces allow for flux conserving, bin averaging from LLC output to lon/lat grid?
The application of this is bin averaging of ECCO-Darwin air-sea CO2 fluxes for
various end users in the carbon community.

Thanks,
Dustin

grid_load error using Matlab 2018b

Hi Gael,

I ran into the following error using Matlab 2018b Update 2 with a fresh copy of gcmfaces that I downloaded yesterday (12 Feb). I got this error both on MacOS (10.14.2) and a Unix HPC system.

I do not get this error with Matlab 2014b.

Error using dir
Characters adjacent to a ** wildcard must be file separators.

Error in rdmds (line 149)
allfiles=dir( sprintf('%s*.meta',fname) );

Error in rdmds2gcmfaces (line 11)
v0=rdmds(varargin{1:end});

Error in grid_load (line 130)
      eval(['mygrid.' list0{iFld} '=rdmds2gcmfaces([mygrid.dirGrid
      ''' list0{iFld} '*'']);']);

Error in animate_subducted_tracer_cuts (line 33)
grid_load(gloc,5,'compact',1);

can't use read_nctiles function in Matlab

I have learned that I must use read_nctiles function to read the ECCO nctiles data. Unfortunately, Matlab keeps reporting errors when running the basic function: " errors read_nctiles (line 172) fldTile=netcdf.getVar(nc,vv,start,count)", and reminds me that the "START" parameter must be 5 numbers but not 4 when call netcdflib function. For example, when I want to read THETA_199201.nc data, ncdisp('THETA_199201.nc') shows the THETA is a 5-D data, but the "start" parameter of read_nctiles seems to be 4 but not 5. I am a rookie to the gcmfaces function toolbox, and I really want to know why this happens.