Artificial intelligence can be used to predict river discharge and warn of potential flooding

Researchers have created a method that uses artificial intelligence to more accurately predict short-term river discharge using historical data from two hydrometric stations on the Ottawa River along with other weather-based parameters. They built on an existing type of algorithm called group method of data handling, which constructs predictive models by sorting and combining data into groups. The models are computed in different combinations repeatedly until the best and most reliable data combination is identified.