When satellites and AI monitor the water: Who gets to shape the data?
KTH researcher Luigia Brandimarte and colleagues recently published "Towards equitable hydrological data in the age of Earth observations and AI" in Nature Water. The authors argue that the future of water science should not be driven by technology alone. Instead, equitable hydrological data require fair access to information, meaningful participation in the production and interpretation of data, and recognition of diverse forms of knowledge, including Indigenous and community-based expertise.
Earth is facing growing water challenges, from droughts and floods to declining freshwater resources. To understand these changes, scientists increasingly rely on satellites and artificial intelligence, technologies that can monitor rivers, lakes, wetlands, and rainfall across the globe, providing information at a scale that was impossible just a few decades ago. But more data does not automatically mean better or fairer decisions.
Most satellite systems and the computing infrastructure needed to process their data are controlled by a small number of technologically advanced countries. Although many satellite datasets are freely available, using them effectively requires expertise, powerful computational infrastructure, and reliable data access. As a result, countries and communities with fewer resources often depend on data and technologies developed elsewhere. This also leaves them vulnerable to shifting priorities and funding conditions.
Water data are never completely neutral. Decisions about what to measure, how to measure it, and which information is considered reliable are shaped by social and political choices.
Writes the researchers: “Satellite images and AI can provide detailed representations of water balances, but they are less able to capture how social-hydrological processes unfold locally and how they are experienced, interpreted, and managed by communities on the ground.” They highlight an example from the Sundarbans, the world’s largest mangrove delta, shared by India and Bangladesh. Satellites can detect coastline changes and saltwater intrusion. Yet, they often fail to capture small-scale flooding patterns, rapidly shifting tidal channels, or changes in water quality that directly affect livelihoods. Local communities, however, experience these changes every day.
Researchers working in the Sundarbans have shown that involving residents in data collection can fill these gaps. Community members measure water salinity, record flooding events, and track environmental changes that formal monitoring systems may overlook. Their observations complement satellite data and help create a more complete picture of local water conditions.
The researchers write that “achieving equitable hydrological data requires concrete changes in how data infrastructures are built and governed towards a practice that is inclusive, participatory, and transparent. Communities, Indigenous groups, practitioners, and decision-makers should be able not only to access data, but also to interrogate, contribute to, and shape the knowledge systems that inform water decisions.”
If we combine new technology with local and experience-based knowledge, AI and ground observations can support a more sustainable water future. But without changing how data is produced and who shapes it, the data revolution may reinforce existing inequalities.
Read the article in Nature Water .