This section focuses on the human dimension of flooding, detailing the social, economic, and environmental impacts.
AI-driven tools, such as satellite imagery analysis and flood prediction models, can be used to track and forecast the impact of floods on vulnerable communities.
Activity: Play the video to find out more about using Sentinel 2 imagery to analyse flooding in the Thessaly region of Greece:
AI–powered tools for recognising flood events
AI-driven tools are increasingly used to recognise and assess flood events through various data sources such as river discharge data, photos of affected areas, and satellite imagery, helping make cities more resilient to flooding.
Activity: Play the video to find out more about Ai and flood management
Platforms provide access to satellite images for flood monitoring, with users able to access and analyse pre- and post-flood conditions. Many of them, like Google Earth Engine and Sentinel Hub are freely available, though some will require registration and coding skills.
<< Flood Risk
The European Commission’s Joint Research Centre (JRC) has used Google Earth Engine to develop high-resolution maps of global surface water occurrence, change, seasonality, recurrence, and transitions.
Their research analyses Landsat images collected over the past 30 years to identify both permanent and seasonal water bodies. This is vital for ensuring water security globally for agriculture, industry, and human consumption. Find out more
The Copernicus Browser is a service provided by the Sentinel Hub. It as a central hub for accessing, exploring and utilisingthe vast amounts of Earth observation and environmental data .
Activity: Try the Copernicus Browser, which can be used anonymously< to generate satellite imagery..
Image source: Copernicus
AI models can be trained on images of flood-affected infrastructure, houses, and environmental elements to detect changes over time.
Google Teachable Machine is one of these tools, characterised by its free, easy-to-use character that allows users to train machine learning models without requiring any prior coding experience. Users can upload their own datasets such as sets of images and teach the machine to recognise specific patterns.
Activity: Play the introductory video to Google Teachable Machine and explore the opportunities it provides for training AI models.
Explore the example lesson using Google Teachable Machine to detect flood photos. Consider the possible ethical issues when using photos or sounds to train an AI model.