Our Platform
Harmful Algal Bloom and Water Quality Forecasts
Check out our harmful algal bloom dashboard for Detroit Lake in Oregon
ClearWater is the platform that synthesizes, cleans and formats all relevant data; custom machine learning algorithms then create accurate forecasts and clear visualizations provide easily interpretable and actionable results.
QUANTIFY
Small, Big, Dirty or Clean: All Data
We scrape, parse and format all data that is relevant to harmful algal blooms. These data can be downloaded by our users for a variety of applications (e.g., GIS). In addition, we create online interactive visualizations for these data, so that users can explore patterns and trends. These data include:
-
weather station,
-
stream gauge,
-
water-sample data and
-
satellite imagery
Our customer HAB prediction algorithms leverage all these available data to make the as accurate as possible forecasts.
PREDICT
Accurate and Resilient Forecasts
The data we collect are used primarily to train custom machine learning algorithms that estimate the future concentration of specific species of (harmful) algae and the toxins they produce. The forecasts target multiple operationally relevant time-horizons, e.g., 10 day look-ahead, as well as long-term seasonal forecasts. ClearWater Analytica was the first to provide HAB forecasts, with an accuracy of >85% depending on the location. Our forecasts are also resilient to data drop-outs, being able to provide accurate predictions even with missing data.
LEARN
HAB Driver Identification
Weather and climate risks are an immediate threat to the provision of clean and drinkable water. Be on the forefront of protecting your key source waters from the impacts of extreme weather events and climate change through insights about your major climate risks, and how risk mitigation strategies can be optimized. ClearWater will quickly provide:
-
An in-depth evaluation of your source water's risks based on our machine learning models
-
Leverage the wealth of peer-reviewed scientific literature and expert opinion
-
Deliver customized risk assessment and mitigation strategies
Publications
White papers, reports and presentations
ClearWater
A short description of our goals and aspirations, as well as the approaches that we use to develop accurate predictions of HABs
ClearWater
An overview of the early HAB prediction system created for the City of Salem in 2019.
Detroit Lake HAB Prediction
Phase 3 Presentation (Fall 2020)
The initial machine learning algorithms for predicting HABs in Detroit Lake were based on Bayesian Model Averaging.
Our Deployment
