Issues with Barents2.5-EPS?

[vic1309]

Hei,

I wonder if anyone has recently faced issues using Barents2.5 -EPS as input in Opendrift. My OD version was running well for files prior 2023, but they can't read now inputs from the model. It looks like wind ensemble is read, but not the ocean currents. This is my code:

import xarray as xr
import matplotlib.pyplot as plt 
import trajan as ta
import numpy as np
import pandas as pd
import cf_xarray as _
import cftime
import pyproj
from datetime import datetime, timedelta
import datetime
from opendrift.readers import reader_netCDF_CF_generic
from opendrift.models.oceandrift import OceanDrift

ds = xr.open_dataset('/lustre/storeB/project/fou/hi/barents_eps/eps/barents_eps_20240310T00Z.nc', engine='netcdf4')

gm = ds.cf['grid_mapping']
beps_crs = pyproj.CRS.from_cf(gm.attrs)
tes = gm.attrs['proj4']

Ds = ds[['u_eastward', 'v_northward', 'Uwind', 'Vwind']]

seeding_time = pd.to_datetime(Ds.time[0].values)

# Start Opendrift
o = OceanDrift(loglevel=0) 

reader_eps = reader_netCDF_CF_generic.Reader(Ds, proj4=tes)

# WAM Model
reader_wave = reader_netCDF_CF_generic.Reader('https://thredds.met.no/thredds/dodsC/sea/mywavewam4/mywavewam4_be')
o.add_reader(reader_wave, variables=['sea_surface_wave_stokes_drift_x_velocity', 'sea_surface_wave_stokes_drift_y_velocity'])

# Workaround to handle EPS products
o.add_reader(reader_eps, variables=['eastward_sea_water_velocity', 'northward_sea_water_velocity', 'x_wind', 'y_wind'])

# Seeding location
seed_lon = 10
seed_lat = 75

o.seed_elements(lon=seed_lon, lat=seed_lat, radius=1000, number=480,
    time=seeding_time, wind_drift_factor=0.02)

# ----- Define processes ----- # 
o.set_config('drift:stokes_drift',  True)
o.set_config('drift:vertical_mixing', False)
o.set_config('drift:vertical_advection', False)

# ----- Perform simulation ----- #
save_path = 'save_path'
o.run(duration=datetime.timedelta(hours=72), time_step=3600, time_step_output=3600, outfile=save_path+'file.nc')

I don't get any errors, but x_sea_water_velocity falls back on 0.

[knutfrode]

Ola, Mr. vic1309!

Yes, the Barents-files changed format so older and newer files are different.
However, recent version of OpenDrift (i.e. reader_netCDF_CF_generic) should handle both, I believe.

Does your log give any indication of why your current becomes 0?