save multiple files opened using mfdataset as a single one

[gkb999]

Hi all,
I use xarray extensively for my research.
However, for clean netcdf files, I could always manipulate directly without having to save them.
However, my current netcdf files, doesn't have time coordinate I suppose.

It looks like the following: (screenshot attached)

here the time variable although it is showing as 3280. this is how ds.time looks like:

For reference: it is a combination of daily data from a particular year (something like say: 2012-2021)
Can I change the time coordinates to the "datetimens" or any standard format that xarray can easily manipulate for further processing?

Also, the x and y cordinates are in 'm', I need them in lat-lon degrees.

Any help on this will be highly appreciated.

[kmuehlbauer]

For the time-problem, you would need to add units (and calendar) attribute. I've tried to prepare the DataArray/Dataset to look like yours.

import xarray as xr
import numpy as np
da = xr.DataArray(np.arange(10), name="time", dims=["time",])
ds = da.to_dataset()
print(ds.time)
ds.time.attrs["units"] = "days since 2012-01-01"
ds.time.attrs["calendar"] = "proleptic_gregorian"
ds = xr.decode_cf(ds)
print(ds)

<xarray.DataArray 'time' (time: 10)>
array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
Dimensions without coordinates: time
<xarray.Dataset>
Dimensions:  (time: 10)
Coordinates:
  * time     (time) datetime64[ns] 2012-01-01 2012-01-02 ... 2012-01-10
Data variables:
    *empty*

Update: Please also refer to the documentation on Creating datetime64 data.

[gkb999]

Awesome, that exactly did what I wanted.
I still think the data doesn't add leap years, but that is still fine, I can ignore those leap years.
How do I add existing variables and the time array?

The data variables are empty. (I'm sorry this could be quite basic, but am new to these kind of analysis)
Many Many thanks in advance already

[kmuehlbauer]

Xarray has a very neat documentation which is continuously improved. You'll find answers to these questions in the user guide and for the specific question in Data Structures

Here are two versions how this might be achieved to get you on the track:

ds = ds.assign(var1=other_ds.var1)
# or
ds["var1"] = other_ds.var1

[kmuehlbauer]

For the projection problem we would need more information on the projection. It might be some polar_stereographic-projection as this is a variable on the dataset in question.

If you need the lat/lon just for plotting, this should work if you use the needed parameters for your projection (if it is polar stereographic):

import cartopy.crs as ccrs

map_proj = ccrs.Stereographic(
    true_scale_latitude=60.0, central_latitude=90.0, central_longitude=10.0
)
ds.z.isel(time=0).plot(subplot_kws=dict(projection=map_proj))

If you need lat/lon as coordinates (then 2D!) there are several packages out there for coordinate transformation (GDAL, tba.).

[gkb999]

This worked too.
Thanks a lot genius :)

[gkb999]

For the projection problem we would need more information on the projection. It might be some polar_stereographic-projection as this is a variable on the dataset in question.

If you need the lat/lon just for plotting, this should work if you use the needed parameters for your projection (if it is polar stereographic):

import cartopy.crs as ccrs

map_proj = ccrs.Stereographic(
    true_scale_latitude=60.0, central_latitude=90.0, central_longitude=10.0
)
ds.z.isel(time=0).plot(subplot_kws=dict(projection=map_proj))

If you need lat/lon as coordinates (then 2D!) there are several packages out there for coordinate transformation (GDAL, tba.).

Just curious why the plot is displayed with a warning:

Local\Temp\ipykernel_6472\3827695759.py:3: UserWarning: "true_scale_latitude" parameter is only used for polar stereographic projections. Consider the use of "scale_factor" instead. map_proj = ccrs.Stereographic(.

[kmuehlbauer]

Glad it worked. But make sure you use the correct parameters (true_scale_latitude, central_latitude, central_longitude). The example values are for a specific projection used by the German Weather Service.

[kmuehlbauer]

Just curious why the plot is displayed with a warning:

Local\Temp\ipykernel_6472\3827695759.py:3: UserWarning: "true_scale_latitude" parameter is only used for polar stereographic projections. Consider the use of "scale_factor" instead. map_proj = ccrs.Stereographic(.

That's maybe due to using some of my aged examples. Probably cartopy evolved and we have to use/adapt the projection. I'll leave this for you to explore.

[gkb999]

Yeah, I changed to south pole, I noticed it is for north. That's really quick and extremely helpful. :) *Girija Kalyani BuradaDoctoral StudentSchool of Geography, Environment and Earth SciencesVictoria University of WellingtonNew Zealand*

…

On Fri, May 12, 2023 at 7:48 PM Kai Mühlbauer ***@***.***> wrote: Glad it worked. But make sure you use the correct parameters ( true_scale_latitude, central_latitude, central_longitude). The example values are for a specific projection used by the German Weather Service. — Reply to this email directly, view it on GitHub <#7839 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABWDHEEXW3GLKBKR7CAAWZ3XFXTONANCNFSM6AAAAAAX6677RI> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[gkb999]

Oh yeah, fine :) This is sufficient . Thanks a lot again :) *Girija Kalyani BuradaDoctoral StudentSchool of Geography, Environment and Earth SciencesVictoria University of WellingtonNew Zealand*

…

On Fri, May 12, 2023 at 7:50 PM Kai Mühlbauer ***@***.***> wrote: Just curious why the plot is displayed with a warning: Local\Temp\ipykernel_6472\3827695759.py:3: UserWarning: "true_scale_latitude" parameter is only used for polar stereographic projections. Consider the use of "scale_factor" instead. map_proj = ccrs.Stereographic(. That's maybe due to using some of my aged examples. Probably cartopy evolved and we have to use/adapt the projection. I'll leave this for you to explore. — Reply to this email directly, view it on GitHub <#7839 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABWDHEC6NZ56RMG4NFF3WLDXFXTUHANCNFSM6AAAAAAX6677RI> . You are receiving this because you authored the thread.Message ID: ***@***.***>