Installation Guide#

Getting started#

HydroMT-Wflow is a model plugin for HydroMT, extending its core functionalities with Wflow-specific components and workflows. It can be installed as a standalone package or alongside other HydroMT model plugins (e.g. HydroMT-SFINCS, HydroMT-Fiat). We recommend installing HydroMT-Wflow in a dedicated Python environment to ensure dependency consistency.

Prerequisite: Python installation#

You will need Python 3.11 or newer and a package/environment manager such as pip, conda, mamba or uv. These tools simplify installing packages and managing isolated environments.

If you do not yet have one installed, we recommend either:

Both conda variants come preconfigured with the conda-forge channel, which provides free and open packages used by HydroMT.

Installing HydroMT-Wflow#

HydroMT-Wflow is available from both PyPI and conda-forge. The simplest and most flexible approach is to install it using pip inside a new environment.

Installation in a conda environment using uv / pip#

We recommend creating a clean environment to avoid dependency conflicts. For example:

$ conda create -n hydromt-wflow uv python=3.13
$ conda activate hydromt-wflow
$ uv pip install hydromt_wflow

This will install HydroMT-Wflow along with HydroMT core and all required dependencies.

To verify the installation, you can list the installed HydroMT plugins:

$(hydromt-wflow) hydromt --plugins
    Model plugins:
        - model (hydromt 1.3.0)
        - wflow_sbm (hydromt_wflow 1.0.0)
        - wflow_sediment (hydromt_wflow 1.0.0)
    Component plugins:
        - ConfigComponent (hydromt 1.3.0)
        - DatasetsComponent (hydromt 1.3.0)
        - GeomsComponent (hydromt 1.3.0)
        - GridComponent (hydromt 1.3.0)
        - MeshComponent (hydromt 1.3.0)
        - SpatialDatasetsComponent (hydromt 1.3.0)
        - TablesComponent (hydromt 1.3.0)
        - VectorComponent (hydromt 1.3.0)
    Driver plugins:
        - dataset_xarray (hydromt 1.3.0)
        - geodataframe_table (hydromt 1.3.0)
        - geodataset_vector (hydromt 1.3.0)
        - geodataset_xarray (hydromt 1.3.0)
        - pandas (hydromt 1.3.0)
        - pyogrio (hydromt 1.3.0)
        - raster_xarray (hydromt 1.3.0)
        - rasterio (hydromt 1.3.0)
    Catalog plugins:
        - deltares_data (hydromt 1.3.0)
        - artifact_data (hydromt 1.3.0)
        - aws_data (hydromt 1.3.0)
        - gcs_cmip6_data (hydromt 1.3.0)
    Uri_resolver plugins:
        - convention (hydromt 1.3.0)
        - raster_tindex (hydromt 1.3.0)

Installing optional dependencies#

HydroMT-Wflow provides several optional dependencies that extend its capabilities, such as additional data sources or hydrological processing functions. You can install these easily using pip’s extras syntax:

$(hydromt-wflow) uv pip install "hydromt_wflow[extra]"

This will install optional packages such as:

gwwapi - provides access to Global Water Watch reservoir datasets.
hydroengine - enables integration with Google Earth Engine.
wradlib - provides radar rainfall processing and interpolation tools.
pyet - adds evapotranspiration computation support.

For a list of all the optional dependency groups and their contents, have a look at the pyproject.toml file. Use hydromt_wflow[full] to install all optional dependencies.

Installing via conda#

HydroMT-Wflow is also available through the conda-forge channel. You can install it directly with:

$ conda create -n hydromt-wflow -c conda-forge hydromt_wflow
$ conda activate hydromt-wflow

Note that some optional dependencies (e.g. gwwapi or hydroengine) are only available through PyPI. You can install them afterwards with pip inside your conda environment:

$(hydromt-wflow) uv pip install "hydromt_wflow[extra]"

Developer installation#

If you want to contribute to HydroMT-Wflow or modify its source code, see the Developer installation guide.

For development work, you can use either a Conda-based setup or Pixi, which provides a fully reproducible project environment. Pixi should be used only in developer installations — not for general users — since it manages dependencies project-locally and is less suited for managing multiple plugins globally.