Using Python to drive Australian water availability forecasting
summary
A Python-based framework named WAFARI generates monthly three-month streamflow forecasts for about 78 sites using a Bayesian statistical core now being replaced by dynamic hydrological models; models are glued with f2py, run from IPython toolkits, and plotted with matplotlib. A separate Hydrological Reference Stations pipeline precomputes images and tables for 221 sites using pandas, HDF5 and SQLAlchemy, serves them as static pages, and reproduces forecasts reliably for operational use.
AI-written summary · about
chapters (15)
- 0:00Start
- 1:01Outline
- 4:26Water Information Role
- 4:53Extended Hydrological Prediction
- 6:07WAFARI: Seasonal streamflow forecasts
- 8:37Features: Automation
- 9:04Features: Bespoke products designed for the SSF service.
- 9:51Challenge: scientific model core
- 10:45Solution: scientific model core
- 11:21Challenge: data management
- 12:06Solution: data management
- 13:21Hydrologic Reference Stations
- 14:12Challenges
- 15:42Solutions
- 18:04What makes our life easier?
related talks
33mPython for Water Forecasting Services
31mHow Python is supporting Australia’s energy transition to 100% renewable
28mDiving into the Deep End: A Python Journey in Water Management
17mGoing with the flow? Apache Airflow for operational-quality scientific workflows
21mAquawatch Data Systems a cloud solution for water quality monitoring and forecasting
26mFrom Data to Decision: Monitoring and Improving Water Networks using AI/ML
22mPython for bridging between researchers and service operators: from CFFI to Jupyterhub
19mMaking Waves with R, Python, and Quarto