io.read_grid for generic NetCDF files?

[adelmoral2]

Hi all!

While trying to open a gridded file with PyArt, that was originally created with LROSE Radx2Grid, I realized that it would only work with PyArt-gridded data.

I ended up working with XARRAY, since we can't turn off those reserved variables.

Would it be possible to work on a version that can also deal with generic NetCDF in the future, using cf-attributes?

Thanks!

[zssherman]

Hi @adelmoral2 What errors are you running into? It should work with generic netcdf files that are in cfradial format, what are the variables called in the dataset you are using?

[adelmoral2]

HI @zssherman this is the error I am encountering:
It is not able to find the "origin_latitude", etc. reserved variables, which I don't know if I can turn off.

These are my variables:

variables:
double time(time) ;
time:standard_name = "time" ;
time:long_name = "Data time" ;
time:units = "seconds since 1970-01-01T00:00:00Z" ;
time:axis = "T" ;
time:bounds = "time_bounds" ;
time:comment = "2023-07-06T11:16:00Z" ;
double start_time(time) ;
start_time:long_name = "start_time" ;
start_time:units = "seconds since 1970-01-01T00:00:00Z" ;
start_time:comment = "2023-07-06T11:10:08Z" ;
double stop_time(time) ;
stop_time:long_name = "stop_time" ;
stop_time:units = "seconds since 1970-01-01T00:00:00Z" ;
stop_time:comment = "2023-07-06T11:16:00Z" ;
double time_bounds(time, bounds) ;
time_bounds:comment = "time_bounds also stored the start and stop times, provided the time variable value lies within the start_time to stop_time interval" ;
time_bounds:units = "seconds since 1970-01-01T00:00:00Z" ;
float x0(x0) ;
x0:standard_name = "longitude" ;
x0:long_name = "longitude" ;
x0:units = "degrees_east" ;
x0:axis = "X" ;
float y0(y0) ;
y0:standard_name = "latitude" ;
y0:long_name = "latitude" ;
y0:units = "degrees_north" ;
y0:axis = "Y" ;
float z0(z0) ;
z0:standard_name = "altitude" ;
z0:long_name = "constant altitude levels" ;
z0:units = "km" ;
z0:positive = "up" ;
z0:axis = "Z" ;
int grid_mapping_0 ;
grid_mapping_0:grid_mapping_name = "latitude_longitude" ;
float DBZ_TOT(time, z0, y0, x0) ;
DBZ_TOT:valid_min = -3.4028e+38f ;
DBZ_TOT:valid_max = 3.4028e+38f ;
DBZ_TOT:_FillValue = -9999.f ;
DBZ_TOT:min_value = -22.46047f ;
DBZ_TOT:max_value = 46.60222f ;
DBZ_TOT:standard_name = "DBZ_TOT" ;
DBZ_TOT:long_name = "DBZ_TOT" ;
DBZ_TOT:units = "dBZ" ;
DBZ_TOT:grid_mapping_name = "latitude_longitude" ;
float DBZ(time, z0, y0, x0) ;
DBZ:valid_min = -3.4028e+38f ;
DBZ:valid_max = 3.4028e+38f ;
DBZ:_FillValue = -9999.f ;
DBZ:min_value = -29.37027f ;
DBZ:max_value = 35.58571f ;
DBZ:standard_name = "DBZ" ;
DBZ:long_name = "DBZ" ;
DBZ:units = "dBZ" ;
DBZ:grid_mapping_name = "latitude_longitude" ;
float VEL(time, z0, y0, x0) ;
VEL:valid_min = -3.4028e+38f ;
VEL:valid_max = 3.4028e+38f ;
VEL:_FillValue = -9999.f ;
VEL:min_value = -20.18224f ;
VEL:max_value = -6.038569f ;
VEL:standard_name = "VEL" ;
VEL:long_name = "VEL" ;
VEL:units = "m/s" ;
VEL:grid_mapping_name = "latitude_longitude" ;

// global attributes:
:Conventions = "CF-1.6" ;
:history = "Sigmet RAW file: Applying azimuth offset: 0, mod time 2025/03/25 22:43:07\nApplying elevation offset: 0, mod time 2025/03/25 22:43:07\n" ;
:institution = "EOL/NCAR" ;
:source = "Sigmet IRIS software" ;
:title = "zar_radar" ;
:comment = "" ;
}

[zssherman]

@adelmoral2 Can you try reading with pyarts aux_io read_radx_grid?

pyart.aux_io.read_radx_grid

[adelmoral2]

@zssherman

I have the same issue:

---

KeyError Traceback (most recent call last)
File ~/miniforge3/envs/pyart_env_n/lib/python3.13/site-packages/pyart/aux_io/radx_grid.py:81, in read_radx_grid(filename, exclude_fields, **kwargs)
80 try:
---> 81 origin_latitude = _ncvar_to_dict(dset.variables["lat0"])
82 # set data single value from Grid info origin from radx file

KeyError: 'lat0'

During handling of the above exception, another exception occurred:

AttributeError Traceback (most recent call last)
Cell In[13], line 1
----> 1 grid=pyart.aux_io.read_radx_grid('ncf_20230706_122559.nc')

File ~/miniforge3/envs/pyart_env_n/lib/python3.13/site-packages/pyart/aux_io/radx_grid.py:100, in read_radx_grid(filename, exclude_fields, **kwargs)
97 except KeyError:
98 origin_latitude = get_metadata("latitude")
99 origin_latitude["data"] = [
--> 100 dset.variables["grid_mapping_0"].latitude_of_projection_origin
101 ]
103 origin_longitude = get_metadata("longitude")
104 try:

File src/netCDF4/_netCDF4.pyx:5030, in netCDF4._netCDF4.Variable.getattr()

File src/netCDF4/_netCDF4.pyx:4752, in netCDF4._netCDF4.Variable.getncattr()

File src/netCDF4/_netCDF4.pyx:1667, in netCDF4._netCDF4._get_att()

File src/netCDF4/_netCDF4.pyx:2164, in netCDF4._netCDF4._ensure_nc_success()

AttributeError: NetCDF: Attribute not found

[zssherman]

@adelmoral2 I'm trying to think of a solution to this. I see the grid Cartesian coords shared above are:

x0:standard_name = "longitude" ;
x0:long_name = "longitude" ;
x0:units = "degrees_east" ;
x0:axis = "X" ;
float y0(y0) ;
y0:standard_name = "latitude" ;
y0:long_name = "latitude" ;
y0:units = "degrees_north" ;
y0:axis = "Y" ;

These should be in meters and not degrees and could also be adding to the issues with the reader.

[RBhupi]

@zssherman Can we add explicit error message?
Is this geographic_to_cartesian going to help make it compatible?