Wflow

Short Name: Short name of model
Wflow
Long Name: Long name of model
wflow_sbm
Model Type: Event or continuous in time; distributed, semi-distributed, or lumped in space
Continuous in time; fully distributed (gridded) in space.
Flood Mechanism: Riverine flood or urban flood or rural flash flood or urban flash flood; rainfall-only or rainfall-snowmelt derived; note wildfire/debris-flow applicability where relevant
Riverine flooding via river routing with 1D floodplains or combined 1D river + 2D overland; also pluvial overland flow.
Usage: Demonstrated application purposes with emphasis on flash flood and riverine flood forecasting; include wildfire/debris use-cases where applicable
Used for riverine/flash-flood forecasting and water resources at scales from small catchments to continental; integrates in Delft-FEWS for operational forecasting.

In general, Wflow supports a wide range of hydrological applications, including:
  • Flood modelling inputs
  • Reservoir inflow calculations
  • Inputs for water quality modelling
  • Sediment inflow assessments
  • Groundwater recharge estimation
Special Features: Unique flood mechanisms, unique usage, additional special simulation purposes
Modular routing (kinematic wave / local-inertial including optional floodplains), glacier & snow modules, BMI interface, FEWS integration.
Background: Description of model history and background
Originated at Deltares (2008+), Python + PCRaster lineage (OpenStreams) → Julia engine from 2021 with active development.
Channel Routing: Available methods for flood wave routing
Kinematic wave (default) and local-inertial options; floodplain routing in 1D or 2D (with 2D overland).
Reservoir Operation: Available methods for reservoir routing and gate control operations if any
Reservoirs/Lakes supported with rating/storage curves and operational parameters (demand, min/max target fractions, max release).
Shortwave Radiation: Available methods for shortwave radiation
Not explicitly solved; Wflow uses potential/reference ET as forcing rather than computing radiation balance.
Longwave Radiation: Available methods for longwave radiation
Same as above.
Precipitation: Available methods for precipitation
Uses gridded precipitation forcing (NetCDF).
Evapotranspiration: Available methods for potential evapotranspiration
The Land Hydrology SBM model assumes the input to be potential reference evapotranspiration. A crop coefficient is used to convert the potential evapotranspiration rate of a reference crop fully covering the soil to the potential evapotranspiration rate of vegetation (natural and agricultural) fully covering the soil.

The crop coefficient is used for a surface completely covered by vegetation and does not include the effect of growing stages of vegetation and soil cover. These effects are handled separately through the use of the canopy gap fraction.
Snowmelt: Available methods for snow accumulation and melt
Temperature-index (degree-day) snow with melt/refreezing (parameters: cfmax, cfr, tt, tti). Optional gravitational transport. Glacier melt (degree-day) supported.
Infiltration: Available methods for infiltration
Infiltration capacity split between compacted and non-compacted areas; actual infiltration limited by remaining unsaturated storage. Optional frozen-soil reduction. Infiltration-excess and saturation-excess runoff occur when capacity or storage are exceeded.
Surface Runoff: Available methods for surface runoff
Overland flow routed by kinematic wave (or 2D local-inertial when enabled).
Interflow: Available methods for interflow
Lateral subsurface flow (kinematic wave across D8 network) or groundwater flow (unconfined aquifer to 4 neighbors).
Percolation: Available methods for percolation
Loss to deep groundwater supported via a (maximum) leakage parameter.
Baseflow: Available methods for baseflow
No clear distinction between baseflow and interflow; see interflow.
Input Data: What types of input data are required for the model
Static maps (DEM, LDD/flow direction, rivers, land-use, soils, gauges) + forcing (P, PET, T; NetCDF); CSV for specific tables (e.g., lakes/reservoirs).
Input Format: What file formats can be used for input data
NetCDF for static & dynamic inputs; TOML settings; CSV for certain tables (e.g., storage/rating curves).
Input Time Interval: What time interval is required for input data, and how are daily values utilized
Time info in TOML is optional; if omitted, model uses timestamps from forcing NetCDF. Supports daily or sub-daily (e.g., hourly) depending on forcing.
Optimization or Calibration: What tools are available for calibration and optimization
No built-in general calibration scripts. Calibration commonly done via manual parameter tuning.
Data Assimilation: Can observed flow and previous forecast flow be used to update the forecast flow
Yes, via OpenDA.
Ensemble: Can ensemble meteorologic forecasts be used in the model
Yes. FEWS supports ensemble meteorology.
Uncertainty: How is uncertainty represented in the outputs
Represented through ensemble simulations and DA spread in FEWS workflows.
Simulation Time Interval: What time interval is used for simulation
Defined in TOML (start, end, step)
Model Output Time-Series: What time-series outputs are available
Time-series of any mapped model variable, including river discharge, overland flow, and precipitation aggregated by area. Variables can be outputted as gridded maps per model timestep, or aggregated by area/catchments. Additional variables can be exported by defining them in the [netcdf.variable] or [csv] sections of the TOML file.
Time-Series Format: What is the file format for output time-series data
NetCDF (gridded), NetCDF (scalar) and CSV.
Model Output Statistics: What types of output statistics are available
Not available, done via post-processing.
Statistics Format: What is the file format for output statistics data
Not available, done via post-processing.
Inventory Platform: Platforms from this inventory in which this model can be integrated
Delft-FEWS (operational forecast platform)
Additional Platform: Additional platforms outside this inventory in which this model can be integrated
Interoperable with MODFLOW, SFINCS (flood), D-Flow FM/1D, DELWAQ, Ribasim etc.
Installation: Difficulty level for installation and configuration
Wflow can be used in two different ways, depending on the required use of the code: If you want to stay up-to-date with the latest version, explore and modify the model code, and write your own Julia scripts around the wflow package, we recommend installing wflow as a Julia package. If you don’t need extra features, but just want to run simulations, a compiled executable version is available. This version includes a single executable, wflow_cli, which allows you to run the model via the command line.
User Education: Education level recommended for users
Users should be comfortable with hydrology + NetCDF/TOML, basic python/CLI. Extensive docs and GMD paper available
Degree of Difficulty: Score from 1 (difficult) to 5 (easy) rating the overall difficulty of use
3/5 (indicative): engine operation is straightforward once inputs are prepared
GIS Support: How much GIS support is included for watershed delineation and parameter estimation
Delination is done using the HydroMT-Wflow plugin (see below). The model files and geometries can be visualized using GIS, but there is no specific GIS tool for delinieation and/or parameter estimation.
Data Preparation: What tools are included for importing and preparing time-series data
The HydroMT-Wflow plugin can be used as a command line application, which provides commands to build, update (including clip) a Wflow model with a single line, or from python to exploit its rich interface.
Land Surface Parameters: Parameters for infiltration, surface runoff, baseflow for each catchment
Soil parameters (conductivity, saturated/residual water contents, Brooks-Corey parameter), vegetation parameters (rooting depth, LAI, crop factors), land cover parameters (fraction paved, fraction open water, roughness)
Parameter Estimation: What tools are included for estimating model parameters from physical data
Initial parameter estimation done via pedo-transfer-functions; Manual tuning common; OpenDA for formal optimization.
Model Calibration: What tools are included for model calibration
Example calibrations exist, but no built‑in auto‑calibration in core Julia engine yet.
Model Verification: What tools are included for model verification
Compare simulated vs observed at gauges (CSV/NetCDF outputs). HydroMT provides notebooks for plotting/analysis of Wflow outputs.
Hardware Requirements: Minimum hardware requirements
4GB of RAM, CPU with 1 core/thread is sufficient, although speedup can be achieved by using more threads
Operating System: Operating system – MS Windows, LINUX
Windows/Linux/macOS
Language of Core Code: Programming language used for the core code (e.g. Fortran, C++, Java). If workflow scripting is supported (e.g. Python) then please specify
Julia
Open Source: Open source or closed source
Open Source
Last Update and Version: Date of latest update and the version number for the release
Jan 07, 2026: v1.0.1
Next Update and Version: Date of next planned update and the version number for the release
Active Development Community: Is there an active developer community with regular updates and new releases?
Active GitHub repos with regular commits/releases and issue tracking: https://github.com/Deltares/Wflow.jl/issues
Platform integration: Platforms in the inventory which integrate this model
Integrates with Delft‑FEWS (operational), OpenDA, and interoperable with MODFLOW/SFINCS/DFM/DELWAQ ecosystems.
Download URL: URL that can be used to download the software
Free to Download and Use: Is the software free to download and use?
Yes
Language of Software Interface: Languages used for the software user interface
Wflow does not have a graphical user interface (GUI). It is operated via command-line interface (CLI) and configuration files. Documentation, code comments, and training materials are available in English, satisfying accessibility requirements for technical users.
Online Support URL: URL that can be used to get online support
Training Material URL (including example data sets): URL that can be used to access training material
Language of Trainings: Languages used for the training material
English
Guidance Material URL (including case studies and benchmarking of performance/speed): URL for case studies and examples of its use. Ideally including benchmarking of performance
Language of Guidance: Languages used for the guidance material
English
References: Reference from scientific journals or publications
Main reference: van Verseveld, W. J., Weerts, A. H., Visser, M., Buitink, J., Imhoff, R. O., Boisgontier, H., Bouaziz, L., Eilander, D., Hegnauer, M., ten Velden, C., and Russell, B., 2024. Wflow_sbm v0.7.3, a spatially distributed hydrological model: from global data to local applications. Geosci. Model Dev., 17, 3199–3234. https://doi.org/10.5194/gmd-17-3199-2024. Please see the list of scientific journals or publications: https://deltares.github.io/Wflow.jl/dev/home/publications.html#peer-reviewed-journal-papers
Owner, Developer: Contact organization for the software. Could be core developer
wflow@deltares.nl (owner and developer) and ali.meshgi@deltares.nl (developer)

Download hydrologic model template