Water Hackers

Turning Uncertainty Into Clear Decision

  • 0 Raised
  • 63 Views
  • 0 Judges

Tags

  • North Macedonia

Categories

  • Challenge #3: Disaster risk monitoring​
  • Challenge #2: Tracking and preventing water pollution​

Description

💎 Idea

AxiosFloodcast is an operational flood forecasting system that combines real-time satellite monitoring with physics-based hydraulic modeling to provide early warning and predictive flood information for hydrometeorological agencies in the Balkans and European regions. The system operates on two integrated pillars: Axios (the automatic flood modeling base) continuously runs HEC-RAS hydrodynamic flood simulations (Rain-on-Grid) forced by hourly ECMWF rainfall forecasts at 9 km resolution (0–72 hours lead time with highest skill, extended to 10 days for advisory alerts), while Floodcast (the real-time monitoring and prediction module) ingests daily soil moisture observations from Copernicus Sentinel-1 satellites (1 km resolution) to dynamically adjust infiltration parameters and compute a Hazard Score which is a probabilistic metric that multiplies forecasted rainfall intensity (compared against user-provided Intensity-Duration-Frequency curves) by soil saturation effects to quantify flood probability and magnitude updated every 6 hours. The Hazard Score is calibrated via hindcast validation to translate raw scores into actionable probabilities (e.g., score 150 = 60% flood probability), enabling authorities to make tiered decisions: probabilities >70% trigger evacuation orders, 50–70% alert preparedness and <50% monitor. Output includes flood extent maps, water depths, velocities, arrival times and hazard probability, allowing decision-makers up to 72 hours advance notice to evacuate populations and deploy resources, while the system's continuous updates provide real-time visibility into whether developing storms are intensifying or subsiding. 


🛰️ EU space technologies

Active Operational Data (Real-time / Daily):

  • ECMWF IFS – Rainfall forecasts (9 km, hourly, every 6 hours)
  • Sentinel-1 (CGLS) – Soil moisture SSM/SWI (1 km, daily)
  • Metop ASCAT – Soil moisture validation (10 km, daily)

Static/Reference Data:

  • EU-DEM – Digital elevation model (25 m)
  • CLMS Land Cover – Land use classification (10–100 m)
  • SoilGrids / EU Soils – Soil properties (250 m)
  • EU Hydro – Basin boundaries (10–100 m)

Fallback/Complementary:

  • ESA CCI (Copernicus Climate Data Store) – Merged soil moisture (25 km, daily, 10–20 day delay)

User-Provided:

  • IDF Curves – Local rainfall frequency analysis
  • Optional: Custom HEC-RAS geometry, land-use maps


 EU Space for Water

We are solving challenges #2: Tracking and preventing water pollution and #3: Disaster risk monitoring.


🤼 Team

  • Aleksandar Ristoski - Project Leader and AI/ML Engineer
  • Maximilian Gligorov - Hydrotechnics Civil Engineer 
  • Anastasija Kotevska - Environmental Technologist
  • Petar Bozhinovski - Mechanical Engineer
  • Georgij Janev - Business Developer
  • Risto Bozhinov - Software Engineer

Attachments