hydroflows 0.1.0

HydroFlows: Automated and reproducible hydro model workflows.

Warning

This is a work in progress! Deltares does not provide any support for this software. The software is provided as is and is not suitable for production use.

Overview

HydroFlows aims to make it easy to create validated workflows using standardized methods and parse these to a workflow engine. In HydroFlows, a workflow consists of methods that are chained together by connecting the file-based output of one method to the input of another. While HydroFlows can also execute the workflows, it is recommended to use a workflow engine to manage the execution of the workflows and to take advantage of the parallelization, scalability, and caching features of these engines. Currently we support Snakemake or engines that support the Common Workflow Language (CWL).

Why HydroFlows?

It can be challenging to create workflows, especially when these should be modular and flexible. With HydroFlows, users can create workflows in a Python script and don’t need to learn a new language or syntax. Using a IDE such as VSCode method in- and outputs can easily be discovered, making it easy to chain methods together in a workflow. Furthermore, method parameters are directly validated at initialization and connections between methods are validated when adding them to the workflow. All these features make it easy to create and maintain workflows compared to other workflow engines.

HydroFlows for flood risk assessments

Currently, the available methods in HydroFlows are focused on flood risk assessments. Methods include the automated setup of physics-based models such as Wflow and SFINCS, statistical analysis, and impact assessments using Delft-FIAT. Many methods build on HydroMT and are backed up by a large stack of state-of-art global datasets to enable rapid assessments anywhere globally. As the workflows are fully automated these can easily be replaced by local data where available. The final outcomes of the HydroFlows flood risk workflows are flood hazard and risk maps and statistics. In addition a FloodAdapt instance can be created from the built models and event sets.

Acknowledgements

This library was created as part of the Horizon Europe UP2030 (Grant Agreement Number 101096405) and InterTwin (Grant Agreement Number 101058386) projects.

License

MIT license, see the LICENSE file for details.