Forbes
A satellite can see a flood through clouds and darkness. It cannot decide whether to close a parish road, move emergency equipment, open a shelter, or send a rescue team. That distinction is increasingly important as governments and emergency managers gain access to sophisticated weather models, artificial intelligence and Earth-observation data. The technical challenge of detecting a flood is becoming easier. The harder problem is translating enormous volumes of data into decisions that people can make quickly, and ensuring those decisions reflect what communities actually need.
A new partnership between Finnish satellite company ICEYE and Louisiana-based The Water Institute is an attempt to bridge those worlds. The companies are pairing ICEYE’s synthetic aperture radar, or SAR, observations with FloodID, a predictive flood-intelligence platform developed in Louisiana for emergency managers. FloodID combines hydrologic modeling, AI, Earth-observation data and cloud computing to anticipate how flooding may develop, while ICEYE can observe actual conditions as an event unfolds.
The model is unusually coherent for disaster technology: predict, observe, translate and act. But a fifth element cannot be supplied from orbit: community experience, institutional relationships, and trust.
Flood Forecasting Needs More Than Better Data
The Water Institute emerged from Louisiana’s experience with hurricanes, coastal loss and other water-related disasters. Founded in 2011, it operates as an independent applied research organization positioned between government, universities, industry and communities. Its mission is not simply to produce science, but to turn it into planning and decision tools. More recently, Louisiana created a Water Innovation Studio in partnership with the Institute to help move water technologies from research through testing and commercialization.
That history matters because FloodID did not begin as a speculative technology looking for a problem. Hugh Roberts, who helped lead the platform’s development at The Water Institute and is now ICEYE’s Head of Predictive Intelligence, said its origins can be traced to emergency managers asking for information they could use while making operational decisions.
“Is it closing roadways? Is it search and rescue? ... Is there a condition that unfolded that we weren’t totally prepared for, but now we have to make some action decisions in short order?” Roberts said.
A 2023 presentation to Louisiana’s Coastal Protection and Restoration Authority shows how explicitly the system was designed around that problem. Its decision-support tools included roadway inundation, search and rescue, response infrastructure and estimates of direct damage. The presentation also described a system built to combine rainfall, streamflow and storm-surge information and translate the modeling into tools tailored to stakeholder needs.
The system has since moved beyond research. The Water Institute says FloodID is being used in emergency operations centers in several hurricane-prone jurisdictions. A 2026 pilot supported by Amazon and AWS has enrolled more than 200 federal, state and local government users.
ICEYE potentially adds another important layer. The company began at Finland’s Aalto University, where founders Rafal Modrzewski and Pekka Laurila worked on making SAR satellites small enough to deploy at scale. ICEYE launched its first sub-100-kilogram SAR satellite in 2018 and subsequently expanded into disaster monitoring for governments and insurers. The new system is intended to create a feedback loop. FloodID predicts where water may go. ICEYE observes where it actually goes. Those observations can then update the forecast as the event evolves, rather than leaving emergency managers with one prediction created before conditions changed.
Community Flood Resilience Depends On Trust And Local Knowledge
Better prediction, however, does not automatically produce community resilience. FloodID is purpose-built primarily for emergency managers rather than as a consumer warning application. Roberts described it as a system for authenticated users who have responsibility for coordinating response and communicating with the public.
That places considerable importance on the institutions between the technology and the resident.
Beaux Jones, president and CEO of The Water Institute, put the issue more simply during our conversation: “We know that the solutions that are hatched closest to the problem are generally the best solutions.”
Research increasingly supports that view. A 2022 study on “smart flood resilience” found that community-scale sensor, mobility, infrastructure and other data can improve predictive monitoring and situational awareness, including understanding which roads may flood and whether residents could lose access to critical facilities such as hospitals.
A more recent study in Gulfport, Mississippi, approached the problem from the other direction. Researchers found that community-generated data could identify neighborhood-level conditions that conventional top-down resilience assessments might overlook. The researchers concluded that mapping efforts conducted with underserved communities could improve both the information available to decision-makers and the procedural equity of resilience planning.
That creates an important distinction between data availability and decision capacity. An emergency manager may receive a sophisticated flood forecast, but still needs communications networks, staff, trusted local partners and knowledge of how particular neighborhoods receive information.
Jones emphasized usability as part of that equation. “These are not CSV files that are getting exported to all these emergency managers,” he said. The goal, he explained, is to give decision-makers information they can understand while making consequential choices in real time. As Roberts acknowledged, “Technology can’t solve everything.”
Why SAR Satellites Are Becoming More Valuable For Flood Monitoring
The business case for SAR starts with a physical advantage: radar does not require daylight or clear skies. Unlike optical satellites, SAR systems send their own radar signals toward Earth and measure what comes back. That lets them collect imagery through clouds and in darkness, conditions that are particularly relevant during hurricanes and severe storms.
A comprehensive review of SAR-based urban flood mapping found that the technology offers significant advantages for flood observation because weather and solar illumination do not prevent collection. But the researchers also identified important limitations. Urban environments can be difficult to interpret because buildings, streets and water create complex radar reflections, and spatial and temporal resolution remain important constraints.
The science continues to advance. A 2026 Journal of Hydrology study developed a region-adaptive deep-learning method for near-real-time SAR flood detection, specifically addressing differences between flooding in urban and open areas. Its results demonstrate both the opportunity and the technical challenge: SAR can support rapid, high-precision flood mapping, but accurately interpreting complex urban environments requires more sophisticated models.
Commercial demand is expanding along with those capabilities. Defense and intelligence remain a major market for SAR, but disaster response, infrastructure monitoring and environmental applications are broadening its civilian use.
Norway offers one example. In June, its Water Resources and Energy Directorate contracted with ICEYE for nationwide SAR-based flood monitoring across mainland Norway and Svalbard. The service is designed to provide near-real-time flood information while supporting hazard mapping, emergency response and longer-term planning. The commercial value, then, is not simply better satellite imagery. It is shrinking the time between an event occurring, understanding what has happened and deciding what to do next.
The Next Challenge For Disaster Technology: From Flood Maps To Decisions
The ICEYE-FloodID partnership will ultimately be more interesting if it is evaluated not by how much data it produces, but by what happens after that data reaches an emergency operations center.
That raises several questions as FloodID expands. Can federal, state and local agencies see and work from consistent information? Will smaller jurisdictions have access after subsidized pilots end? Can community-generated information improve the models and the decisions made from them? And when broadband, electricity or communications fail during the disaster itself, does the intelligence still reach the people who need it?
Those questions do not diminish the technology. They define its next stage.
The emerging disaster-data ecosystem can increasingly tell officials where water is, where it might go and which infrastructure may be in danger. The harder task is building the human and institutional infrastructure that converts that information into action. A satellite may be able to see through the storm. Resilience still depends on what happens on the ground.