Aquawatch Data Systems a cloud solution for water quality monitoring and forecasting
summary
A cloud platform built on AquaWatch Data Systems and CSIRO’s EASI ingests Sentinel, Landsat and hyperspectral scenes, runs acolyte atmospheric correction (a recent run produced ~0.5 PB of corrected data), collocates ~1,000 in‑situ streams with petabytes of EO in an Open Data Cube, and serves notebooks, JupyterHub, Argo/papermill pipelines, Kubernetes, Terraform, Dask, Trino and FROST integrations; pilots in Australia, Malaysia, Vietnam, Italy and the US demonstrate scalable forecasting and a worst‑case cost estimate near $250k/month.
AI-written summary · about
chapters (20)
- 0:00Introduction and team
- 1:25Aquawatch purpose and data flow
- 2:52Platform investment gap in EO
- 4:40Planetary computer concept
- 5:21AWS architecture components
- 6:03Atmospheric correction with ACOLITE
- 7:29Hyperspectral data support
- 8:53JupyterHub and managed notebooks
- 10:19Paper Mill for production workflows
- 11:24Scaling with Dask
- 12:04Scientist workflow example
- 12:49Open standards for data delivery
- 13:56Institute sensing and ETL
- 15:24Hydrospectra sensor details
- 16:33Trino and FROST for data management
- 17:39USGS water sensor integration
- 18:19Algal bloom monitoring portal
- 19:01Scientist feedback and platform impact
- 19:43Collaboration opportunities
- 20:06Q&A session
related talks
16mThe Open Data Cube and working towards a Digital Earth
14mSatellite data monitoring for analytical models
33mPython for Water Forecasting Services
20mUsing R, Python, and Cloud Infrastructure to Battle Aquatic Invasive Species
28mAdvancing the future of space in the cloud
48mAccelerate science, research, and innovation with AWS
21mCloud-Native Data in Practice
24mManaging Heterogeneous Spatial Data with Analysis Ready Cloud Optimized formats