calc.storm_relative_helicity "TypeError: tuple indices must be integers or slices, not tuple"

[ThunderRoad75]

Calling the SRH function for a single point or spatial subset (in this case, 10x10 grid) throws "TypeError: tuple indices must be integers or slices, not tuple". When attempting to operate on the full volume rather than a single point or subset, it throws a MemoryError. The full RAP volume at 13 km is (37, 337, 451) - which doesn't strike me as excessively large.

Response on Twitter was that it may be a problem with Pint and asked me to open an issue. All errors thrown are attached as a screenshot.

Code snippet below uses UCAR's siphon and THREDDS so should be very easily copied, pasted, and run. Just make sure the valid date is within the last ~30 days.

from time import time
from datetime import datetime, timedelta
from siphon.catalog import TDSCatalog
#from netCDF4 import Dataset
from metpy.calc import storm_relative_helicity
from metpy.units import units
import numpy as np

# Use siphon to get the desired model dataset and time, and desired variable from the dataset
model_url = "http://thredds.ucar.edu/thredds/catalog/grib/NCEP/RAP/CONUS_13km/RR_CONUS_13km_20190819_1200.grib2/catalog.html?dataset=grib/NCEP/RAP/CONUS_13km/RR_CONUS_13km_20190915_1200.grib2"

# Create a datetime object for the day and FHR being used
vtime = datetime.strptime('2019091512','%Y%m%d%H')

# Get the data
model = TDSCatalog(model_url)
ds = model.datasets[0]
ncss = ds.subset()
query = ncss.query()
query.accept('netcdf4')
query.time(vtime) # Set to the analysis hour only
query.add_lonlat()
query.variables('U-Component_Storm_Motion_height_above_ground_layer', 'V-Component_Storm_Motion_height_above_ground_layer',
'u-component_of_wind_isobaric', 'v-component_of_wind_isobaric', 'Geopotential_height_isobaric')
data = ncss.get_data(query)

# Select variable(s) to plot
hgt = data.variables['Geopotential_height_isobaric'][0,:,10:20,10:20] * units.meter
uwnd = data.variables['u-component_of_wind_isobaric'][0,:,10:20,10:20] * units.meter / units.second
vwnd = data.variables['v-component_of_wind_isobaric'][0,:,10:20,10:20] * units.meter / units.second
print(np.shape(uwnd))
storm_u = data.variables['U-Component_Storm_Motion_height_above_ground_layer'][0,:,10:20,10:20] * units.meter / units.second
storm_v = data.variables['V-Component_Storm_Motion_height_above_ground_layer'][0,:,10:20,10:20] * units.meter / units.second

# Calculate SRH
SRH = storm_relative_helicity(uwnd, vwnd, hgt, 3000. * units.meter, 0 * units.meter, storm_u, storm_v) 

• Running Anaconda on Windows
• Python 3.7.1
• MetPy v0.10.0

[jthielen]

This looks like the same issue as #997 and several others. Would you be able to update MetPy to v0.10.2, as in @zbruick's suggestion in #1146 (comment) and see if that resolves your issue?

This was an issue with our dependence on pint, which caused this issue with their release of pint 0.9 (#997). We released a fix in MetPy 0.10.2, so if you update MetPy (and ask students to do the same), then this issue should be resolved: pip install --upgrade metpy or conda update metpy.

[zbruick]

Yes, if you upgrade to 0.10.2, this first problem should be resolved. The memory problem is a larger issue with how we do these calculations (see #395, #480). This is on the to-do list in the future, but it is not something we support currently. As a result, you would need to loop through your points that you want to this function to work on.

[ThunderRoad75]

Yup, 0.10.2 fixed this issue for a single point. Thanks folks!