index.qmd

---
title: "Streamflow Forecasting"
---

This is a public entry point for a suite of hydrologic ensemble forecasting modelling tools developed in Australia over the past decade. Most scientific modelling features were authored by the [CSIRO](https://www.csiro.au){.external target="_blank"}. While designed with unique features for _ensemble_ streamflow forecasting, these tools can equally be used in _non-ensemble_ simulation mode for other, "traditional" semi-distributed hydrologic modelling.

# Applications

Some of the modelling tools are used by the Australian Bureau of Meteorology (the Bureau) to provide the [7-day Ensemble Streamflow Forecasts](http://www.bom.gov.au/water/7daystreamflow){.external target="_blank"} service. An overview of the research and outcomes in streamflow forecasting of the research alliance between the Bureau and CSIRO is available [at this page](https://www.csiro.au/en/Research/LWF/Areas/Water/Assessing-water-resources/WIRADA/Research/Water-forecasting){.external target="_blank"}.

More recent versions of the tools are also used for further research, development and applications. The core modelling features are usually native (C++) libraries, accessed seamlessly by user from **Python**, **R** and **Matlab**.

![Ensemble forecasting simulation run from an R script](./img/ens_streamflow_forecast.png "Ensemble forecasting simulation run from an R script")

The [documentation](./documentation.qmd) page has links to sample modelling workflows in `R` and `Python` that hydrologic modellers can browse through to get an idea of the features.

# Installation

Installation packages are available for Windows, Debian/Ubuntu Linux, and MacOSX. They currently cannot _all_ be made publicly downloadable, but access for evaluation or research purposes can be arranged. Contact [David Robertson](https://people-my.csiro.au/r/d/david-robertson) at [david.robertson@csiro.au](mailto:david.robertson@csiro.au) and/or [jean-michel.perraud@csiro.au](mailto:jean-michel.perraud@csiro.au). Then follow the instructions at the [Installation](./installation.qmd) page.

# Architecture

The core of the toolset is written in C++ for execution speed, but in practice users mostly access the features via packages in R, python, or Matlab.

![High-level software stack](./img/architecture_poster.png "Software stack overview")

Besides building upon established third party open source software, some components of our stack are also open source. See [csiro-hydroinformatics](https://github.com/csiro-hydroinformatics){.external target="_blank"} on GitHub, for instance the [uchronia time series handling module](https://github.com/csiro-hydroinformatics/uchronia-time-series){.external target="_blank"}.

The [developer page](./doc/developer.qmd) also contains more detailed technical information on the software stack.

# Acknowledgements

Most of the tools covered by this documentation is the output of research supported by the Water Information Research and Development Alliance between CSIRO and the Australian Bureau of Meteorology.

# Publications

[SWIFT2: High performance software for short-medium term ensemble streamflow forecasting research and operations](https://www.mssanz.org.au/modsim2015/L15/perraud.pdf){.external target="_blank"}

[SWIFT2: Advanced software for continuous ensemble short-term streamflow forecasting ](https://search.informit.com.au/documentSummary;dn=823693257390059;res=IELENG){.external target="_blank"}