py_forecasting_2.py

# list of interesting fields
gfs_variables = ['Temperature_surface','u-component_of_wind_isobaric','v-component_of_wind_isobaric',
            'Total_cloud_cover_entire_atmosphere_Mixed_intervals_Average','Low_cloud_cover_low_cloud_Mixed_intervals_Average',
            'Medium_cloud_cover_middle_cloud_Mixed_intervals_Average','High_cloud_cover_high_cloud_Mixed_intervals_Average',
            'Total_cloud_cover_boundary_layer_cloud_Mixed_intervals_Average','Total_cloud_cover_convective_cloud']

gfs_columns = ['temp_air','wind_speed_u','wind_speed_v','total_clouds',
               'low_clouds','mid_clouds','high_clouds','boundary_clouds','convect_clouds', 'time']

df = pd.DataFrame()

for gfs_variable in gfs_variables:
    query = ncss.query()
    query.lonlat_point(long, lat).vertical_level(100000).time_range(now, now + timedelta(days=forecast_days))
    query.variables(gfs_variable).accept('netcdf')
    data = ncss.get_data(query)
    gfs_col = np.array(data.variables[gfs_variable][:].data)
    df = pd.concat([df, pd.DataFrame(gfs_col)], axis=1)

time = data.variables['time']
dates = num2date(time[:], units=time.units, calendar='standard', 
                 only_use_cftime_datetimes=False, only_use_python_datetimes=True)

df = pd.concat([df, pd.DataFrame(dates.data)], axis=1)
df.columns = gfs_columns

df.set_index('time', inplace=True)