Initial Conditions

Similar to other packages, input of initial conditions follows the structure of MODFLOW and Flopy. Setting the starting heads from the model top is often a good way to go initially. After the model has been run, starting heads can then be updated from the initial model head output to improve convergence on subsequent runs.

Note

With any transient model, an initial steady-state stress period is recommended, regardless of initial conditions. Specification of initial conditions is mostly a consideration for improving model convergence.

Starting heads defaults

Omitting or leaving the ic: block blank triggers one of two default actions:

• If there is no parent model, or a parent model with the default_source_data option turned off, starting heads are set to the model top.

• If there is a parent model with the default_source_data option turned on, starting heads are set from the parent model starting heads.

Specifying starting heads directly

For MODFLOW 6 models, the “Options” and “Griddata” input blocks are represented as sub-blocks within the ic: block. Within the griddata: block, model inputs can be specified directly, as long as they are consistent with the definition of the model grid. For example, if strt: [100., 100.] is specified, the model must be either two layers (for “LAYERED” input; see MODFLOW input instructions) or two cells. An example of specifying a uniform starting head condition for a model of any size:

ic:
  griddata:
    strt: 100.

Or specifying uniform starting heads by layer:

ic:
  griddata:
    strt: [100., 99.]

External files can also be supplied, following the Flopy format:

ic:
  griddata:
    strt:
      - {'filename': 'strt_000.dat'}
      - {'filename': 'strt_001.dat'}

Specifying starting heads from grid-independent sources

Similar to other packages, input specified via a source_data: sub-block can be mapped to the model grid from a variety of sources.

Explicitly setting starting heads to the model top (for example, to override the default behavior with default_source_data from a parent model):

ic:
  source_data:
    strt: 'from_model_top'

Setting starting heads from a binary head save file (pathed relative to the configuration file):

ic:
  source_data:
    strt:
      from_parent:
        binaryfile: '/path/to/pleasant.hds'
        stress_period: 0

Explicitly setting starting heads to the parent model starting heads (for example, to override the default behavior if default_source_data is False)

ic:
  source_data:
    strt: from_parent

Example script for updating starting heads from a previous run

The following example script opens MODFLOW binary head results using Flopy, and writes them to external files that can then be input to the IC or BAS6 Packages (for MODFLOW 6 or MODFLOW-NWT models, respectively). If you followed the Modflow-setup defaults to write external files to an external/ folder, with starting head files named as shown in the script, this will overwrite your previous starting head files, requiring no additional action after running the script.

"""Update model starting heads from previous (initial steady-state) solution
"""
from pathlib import Path

import numpy as np
from flopy.utils import binaryfile as bf

model_ws = Path('.')
headfile = model_ws / 'model.hds'
starting_heads_file_fmt = str(model_ws / 'external/strt_{:03d}.dat')


hdsobj = bf.HeadFile(headfile)
print(f'reading {headfile}...')

initial_ss_heads = hdsobj.get_data(kstpkper=(0, 0))
for per, layer_heads in enumerate(initial_ss_heads):
    outfile = starting_heads_file_fmt.format(per)
    np.savetxt(starting_heads_file_fmt.format(per), layer_heads, fmt='%.2f')
    print(f"updated {outfile}")

Download the file: update_starting_heads_from_previous.py

Note: This script can be run as soon as the initial steady-state stress period is done solving. The current MODFLOW run can then be canceled and MODFLOW restarted, hopefully with improved convergence in the initial period.