CRS Error when clipping NetCDF using a shapefile

Question

When trying to mask a NetCDF (a daily time series of 31 days) with the code below, I get the following error related to the part of writing the CRS to the rio NetCDF data:

CRSError: The WKT could not be parsed. OGR Error code 6

Here is the code I am using (taken from snowman2's answer here: How to mask NetCDF time series data from a shapefile in Python?):

import rioxarray
import xarray
from shapely.geometry import mapping

ERA5_daily = xarray.open_dataarray('ERA5_VIMD_2018_08_Daily.nc')
ERA5_daily.rio.set_spatial_dims(x_dim="longitude", y_dim="latitude", inplace=True)
ERA5_daily.rio.write_crs("epsg:4326", inplace=True)
shapefile = geopandas.read_file('Cluster_34586.shp', crs="epsg:4326")

clipped = ERA5_daily.rio.clip(shapefile.geometry.apply(mapping), shapefile.crs, drop=False)

Does anyone know how to fix it?

Answer 1

Using this code:

import geopandas
import rioxarray
import xarray
from shapely.geometry import mapping

ERA5_daily = xarray.open_dataarray('ERA5_VIMD_2018_08.nc')
ERA5_daily.rio.write_crs("EPSG:4326", inplace=True)
shapefile = geopandas.read_file('Cluster_34586.shp', crs="EPSG:4326")
clipped = ERA5_daily.rio.clip(shapefile.geometry.apply(mapping), shapefile.crs, drop=False)

It works with this setup (from: rioxarray.show_versions()):

rioxarray (0.0.31) deps:
  rasterio: 1.1.5
    xarray: 0.16.0
      GDAL: 3.0.4

Other python deps:
     scipy: 1.5.2
    pyproj: 2.6.1.post1

System:
    python: 3.7.8 | packaged by conda-forge | (default, Jul 31 2020, 02:25:08)  [GCC 7.5.0]

I would recommend upgrading the versions you have on your system.