ogs5py 1.3.0

ogs5py: a python API for OpenGeoSys5

Welcome to ogs5py

Purpose

ogs5py is A python-API for the OpenGeoSys 5 scientific modeling package.

Installation

You can install the latest version with the following command:

pip install ogs5py

Or with conda:

conda install ogs5py

Citation

If you are using ogs5py in your publication please cite our paper:

Müller, S., Zech, A. and Heße, F.: ogs5py: A Python-API for the OpenGeoSys 5 Scientific Modeling Package. Groundwater, 59: 117-122. https://doi.org/10.1111/gwat.13017, 2021.

You can cite the Zenodo code publication of ogs5py by:

Sebastian Müller. GeoStat-Framework/ogs5py. Zenodo. https://doi.org/10.5281/zenodo.2546767

If you want to cite a specific version, have a look at the Zenodo site.

Documentation for ogs5py

You can find the documentation under ogs5py.readthedocs.io.

Further Information

• General homepage: https://www.opengeosys.org/ogs-5
• OGS5 Repository: https://github.com/ufz/ogs5
• Keyword documentation: https://ogs5-keywords.netlify.com
• OGS5 Benchmarks: https://github.com/ufz/ogs5-benchmarks
• ogs5py Benchmarks: https://github.com/GeoStat-Framework/ogs5py_benchmarks

Tutorials and Examples

In the following a simple transient pumping test is simulated on a radial symmetric mesh. The point output at the observation well is plotted afterwards. For more details on this example, see: Tutorial 1

from ogs5py import OGS, specialrange, generate_time
from matplotlib import pyplot as plt

# discretization and parameters
time = specialrange(0, 3600, 50, typ="cub")
rad = specialrange(0, 1000, 100, typ="cub")
obs = rad[21]
angles = 32
storage = 1e-3
transmissivity = 1e-4
rate = -1e-3
# model setup
model = OGS(task_root="pump_test", task_id="model")
# generate a radial mesh and geometry ("boundary" polyline)
model.msh.generate("radial", dim=2, rad=rad, angles=angles)
model.gli.generate("radial", dim=2, rad_out=rad[-1], angles=angles)
model.gli.add_points([0.0, 0.0, 0.0], "pwell")
model.gli.add_points([obs, 0.0, 0.0], "owell")
model.bc.add_block(  # boundary condition
    PCS_TYPE="GROUNDWATER_FLOW",
    PRIMARY_VARIABLE="HEAD",
    GEO_TYPE=["POLYLINE", "boundary"],
    DIS_TYPE=["CONSTANT", 0.0],
)
model.st.add_block(  # source term
    PCS_TYPE="GROUNDWATER_FLOW",
    PRIMARY_VARIABLE="HEAD",
    GEO_TYPE=["POINT", "pwell"],
    DIS_TYPE=["CONSTANT_NEUMANN", rate],
)
model.ic.add_block(  # initial condition
    PCS_TYPE="GROUNDWATER_FLOW",
    PRIMARY_VARIABLE="HEAD",
    GEO_TYPE="DOMAIN",
    DIS_TYPE=["CONSTANT", 0.0],
)
model.mmp.add_block(  # medium properties
    GEOMETRY_DIMENSION=2,
    STORAGE=[1, storage],
    PERMEABILITY_TENSOR=["ISOTROPIC", transmissivity],
)
model.num.add_block(  # numerical solver
    PCS_TYPE="GROUNDWATER_FLOW",
    LINEAR_SOLVER=[2, 5, 1e-14, 1000, 1.0, 100, 4],
)
model.out.add_block(  # point observation
    PCS_TYPE="GROUNDWATER_FLOW",
    NOD_VALUES="HEAD",
    GEO_TYPE=["POINT", "owell"],
    DAT_TYPE="TECPLOT",
)
model.pcs.add_block(  # set the process type
    PCS_TYPE="GROUNDWATER_FLOW", NUM_TYPE="NEW"
)
model.tim.add_block(  # set the timesteps
    PCS_TYPE="GROUNDWATER_FLOW",
    **generate_time(time)
)
model.write_input()
model.run_model()

OGS5 executable

To obtain an OGS5 executable, ogs5py brings a download routine:

from ogs5py import download_ogs
download_ogs()

Then a executable is stored in the ogs5py config path and will be called when a model is run.

You can pass a version statement to the download_ogs routine, to obtain a specific version (5.7, 5.7.1 (win only) and 5.8). For OGS 5.7 there are executables for Windows/Linux and MacOS. For "5.8" there are no MacOS pre-builds.

If you have compiled your own OGS5 version, you can add your executable to the ogs5py config path with:

from ogs5py import add_exe
add_exe("path/to/your/ogs/exe")

Otherwise you need to specify the path to the executable within the run command:

model.run_model(ogs_exe="path/to/ogs")

Requirements:

• NumPy >= 1.14.5
• Pandas >= 0.23.2
• meshio >= 4
• lxml >= 4
• pexpect >= 4
• vtk >= 9

Contact

You can contact us via [e-mail].

License

MIT © 2018-2023 (inspired by Falk Hesse and Miao Jing)

This project is based on OGSPY.