AquaGraph_cluj

💎 Idea

Romania's rivers carry pollution that goes undetected for days - by the time it's noticed, ecosystems are damaged and communities are at risk. AquaGraph solves this by combining satellite imagery with a graph model of Romania's river network to monitor water pollution in near real-time. Users get an interactive map, latest news related to water environments and a community hub for cleanup campaigns.

🛰️ EU space technologies

We use Sentinel-1 radar imagery to detect possible oil pollution on water surfaces. Oil slicks can appear as unusually dark and smooth areas because they change the surface roughness of the water. This gives us a useful signal even when clouds or lack of daylight make optical monitoring difficult.

We use Sentinel-2 multispectral imagery to observe visible and spectral changes in water quality. From this, we extract pollution indicators such as water colour changes, turbidity, suspended sediments, algal bloom signals and abnormal spectral index values. These signals help AquaGraph classify river segments as clean, suspicious or polluted.

To understand how rivers are connected, we use the EU-Hydro River Network Database from the Copernicus Land Monitoring Service. This gives us the structure of the river system: river segments, drainage lines, nodes, water bodies and catchment-related information. AquaGraph uses this data to model Romania’s rivers as a graph, where each segment is connected to its upstream and downstream neighbours.

To estimate how pollution could move, we also use GloFAS river discharge and forecast data from the Copernicus Emergency Management Service. Flow information helps us understand the likely direction and speed of pollution spreading downstream.

Together, these data sources allow AquaGraph to do more than simply show pollution on a map. By combining Sentinel-1 oil detection, Sentinel-2 water-quality indicators, EU-Hydro river connectivity and GloFAS flow data, we can identify where pollution is visible, where it may have entered the river system, and where it is likely to travel next.

 🌊EU Space for Water

AquaGraph addresses Challenge 2 (Tracking and preventing water pollution) by providing continuous, satellite-derived pollution tracking across all major Romanian river segments. Unlike point-based sampling, our approach covers the entire hydrological network. By modelling rivers as a connected graph, we show not just where pollution occurs but where it will reach — enabling authorities, industries, and communities to act before damage compounds.

🤼 Team

AquaGraph is developed by six students: three from Babes-Bolyai University, Faculty of Mathematics and Computer Science, and three from the Technical University of Cluj-Napoca, Faculty of Automation and Computer Science. The team is split between frontend, backend, and satellite data processing, with each member contributing to a specific part of the prototype.

Sturza Carina [Frontend Developer]: Carina is responsible for building the user-facing part of the platform. She works on turning the project’s data and features into a clear, usable interface.

Sigmirean Andrada [Frontend Developer]: Andrada works on the visual and interaction side of the application. She focuses on making the platform intuitive, structured, and easy to navigate.

Tîrnăvean Denis  [Backend Developer]: Denis handles the backend development of the platform. He works on the data flow, APIs, and the technical logic that connects the application components.

Carla-Maria Barastean  [Satellite Data Processing]: Carla processes Sentinel-1 radar data for detecting possible oil pollution signals. Her work focuses on extracting relevant information from radar imagery and preparing it for use in the system.

Doru-Calin Lese [Satellite Data Processing]: Doru works on satellite data processing and analysis, supporting the transformation of Earth observation data into usable outputs for the prototype. He also contributes to validating and organizing the processed data.

Tiffany-Monica Vatafu  [Satellite Data Processing]: Monica works with Sentinel-2 multispectral data. She focuses on extracting water-quality indicators from optical satellite imagery and preparing them for integration into AquaGraph.

REPOSITORY: https://github.com/Tifym7/AquaGraph