External files for list-type data in MODFLOW-2005 models

[aleaf]

Hi, does anyone know how to specify an external file for the input of list-type data, for example, for the WEL Package? With the MODFLOW-6 part of Flopy, I can use the code below to write well package data to an external file and then supply that as input. This is convenient for working with the data outside of flopy (both in generating the initial version, and subsequently in parameter estimation or uncertainty quantification).

import pandas as pd
import flopy
mf6 = flopy.mf6
fm = flopy.modflow

# set up simulation
sim = mf6.MFSimulation(sim_name='testsim',
                   sim_ws='tmp')
tdis_rc = [(6.0, 2, 1.0), (6.0, 3, 1.0)]
tdis = mf6.ModflowTdis(sim, time_units='DAYS', nper=2,
                          perioddata=tdis_rc)
ims_package = mf6.ModflowIms(sim)
# create model instance
model = mf6.ModflowGwf(sim, modelname='testmodel')
dis = mf6.ModflowGwfdis(model, length_units='FEET', nlay=1,
                                     nrow=1, ncol=10, delr=500.0,
                                     delc=500.0,
                                     top=100.0, botm=50.0)

# make wel package data and write the external file
wel_data = pd.DataFrame([(0,11,2, -50.0)], 
             columns=['#k', 'i', 'j', 'q'])
wel_data.to_csv('tmp/wel_000.dat', index=False, sep=' ')

# create the wel package instance from the external file
wel = flopy.mf6.ModflowGwfwel(model, pname='wel', print_input=True, print_flows=True,
                              maxbound=1,
                              stress_period_data={0: {'filename': 'wel_000.dat'}})
# access the stress period data
wel.stress_period_data.array

result:

[rec.array([((-1, 10, 1), -50.)],
           dtype=[('cellid', 'O'), ('q', '<f8')]),
 None]

But when I try doing the same thing in a MODFLOW-NWT context, for example, it doesn't seem like the external file input is being properly read.

nwt_model = fm.Modflow('nwt_testmodel')
dis = fm.ModflowDis(nwt_model, nlay=1, nrow=1, ncol=10, delr=500.0,
                    delc=500.0, top=100.0, botm=50.0)
wel = fm.ModflowWel(nwt_model, stress_period_data={0: 'wel_000.dat'})
wel.stress_period_data.array
---------------------------------------------------------------------------
Exception                                 Traceback (most recent call last)
<ipython-input-20-440aa0c07f28> in <module>
----> 1 wel.stress_period_data.array

~/Documents/GitHub/flopy/flopy/utils/util_list.py in array(self)
   1221     @property
   1222     def array(self):
-> 1223         return self.masked_4D_arrays
   1224 
   1225     @classmethod

~/Documents/GitHub/flopy/flopy/utils/util_list.py in masked_4D_arrays(self)
   1176     def masked_4D_arrays(self):
   1177         # get the first kper
-> 1178         arrays = self.to_array(kper=0, mask=True)
   1179 
   1180         # initialize these big arrays

~/Documents/GitHub/flopy/flopy/utils/util_list.py in to_array(self, kper, mask)
   1139                 return arrays
   1140             else:
-> 1141                 raise Exception("MfList: something bad happened")
   1142 
   1143         for name, arr in arrays.items():

Exception: MfList: something bad happened

Trying the MODFLOW-6 style input doesn't work either:

wel = fm.ModflowWel(nwt_model, stress_period_data={0: {'filename': 'wel_000.dat'}})
---------------------------------------------------------------------------
Exception                                 Traceback (most recent call last)
<ipython-input-16-e93205e6bc63> in <module>
----> 1 wel = fm.ModflowWel(nwt_model, stress_period_data={0: {'filename': 'wel_000.dat'}})

~/Documents/GitHub/flopy/flopy/modflow/mfwel.py in __init__(self, model, ipakcb, stress_period_data, dtype, extension, options, binary, unitnumber, filenames)
    262 
    263         # initialize MfList
--> 264         self.stress_period_data = MfList(
    265             self, stress_period_data, binary=binary
    266         )

~/Documents/GitHub/flopy/flopy/utils/util_list.py in __init__(self, package, data, dtype, model, list_free_format, binary)
     95         self.__data = {}
     96         if data is not None:
---> 97             self.__cast_data(data)
     98         self.__df = None
     99         if list_free_format is None:

~/Documents/GitHub/flopy/flopy/utils/util_list.py in __cast_data(self, data)
    374                     self.__vtype[kper] = None
    375                 else:
--> 376                     raise Exception(
    377                         "MfList error: unsupported data type: "
    378                         + str(type(d))

Exception: MfList error: unsupported data type: <class 'dict'> at kper 0

Is there a way to do this? I couldn't find anything in the examples. Thanks in advance.

[jdhughes-dev]

@jtwhite79 Is this supported for MODFLOW-2005.

@spaulins-usgs how would it be done for MODFLOW-6.

@aleaf If it is not currently supported for pre-MODFLOW 6 then it is unlikely that we would prioritize this enhancement. So we would need help from the community (possibly you).

[spaulins-usgs]

@jdhughes-usgs This is currently done for MODFLOW-6 by allowing the user to use a dictionary to define any list or array data. If the dictionary contains a key named "filename", the contents of that key is used as a relative or absolute path to the data file. If the dictionary also has a key named "data", flopy will create the datafile and place the contents of the "data" key in that file. Otherwise, flopy will look for an existing data file at that path.