Description
-Idea-
Our idea is to develop a system for early prediction of floods and droughts using satellite data and machine learning.
Water-related disasters such as floods and droughts are becoming more frequent and harder to predict due to climate change. Traditional monitoring methods are often slow, localized, or limited in coverage.
To address this, we use satellite data to continuously monitor land conditions and detect patterns that indicate potential risk. By applying machine learning models, we transform these patterns into a risk score (0–1), which allows us to identify high-risk areas in advance and support early warning systems.
The system is currently developed and tested for the territory of North Macedonia, but it is designed to be easily scalable and extendable to other regions in the future due to its reliance on globally available satellite data.
-EU Space Technologies-
We use data from the Copernicus programme, specifically:
- Sentinel-1 (SAR radar data) → used for flood detection
(VV and VH signals capture surface structure, moisture, and water presence) - Sentinel-2 (optical data) → used for drought detection
(NDVI and NDWI indices represent vegetation health and water content)
-EU Space for Water-
Our project addresses the challenge of water monitoring and risk management.
We contribute by:
- detecting flood-prone areas using SAR-based moisture signals
- identifying drought conditions using vegetation and water stress indicators
- generating early warning signals for both extremes
This directly supports better water resource management, improves preparedness for natural disasters, and helps reduce their impact on communities, agriculture, and infrastructure.
-Team-
We are a team combining frontend, backend, and machine learning development.
- Elena Hristoska and Marija Ignatoska –Machine Learning Engineers (Python)
Responsible for data processing, feature engineering, and development of predictive models using XGBoost. - Marija Bundoska and Mihaela Hristoska– Frontend Developer (HTML, CSS, JavaScript)
Responsible for designing and implementing the user interface, including the interactive map and visualization of risk results. - Hristijan Ilijoski– Backend & API Developer
Responsible for connecting the frontend and backend through APIs and ensuring smooth data flow between the system components.