Broad coverage with high-resolution predictions where they matter most. Understand flood risk days in advance.
Aqunia Flood is an early warning system that provides real-time forecasts of flood depth and inundation extent several days in advance across large river basins. It supports disaster response and decision-making for governments and infrastructure operators with reliable predictive information.
Current flood warning systems have three key limitations.
Aqunia Flood combines satellite data and AI to deliver high-accuracy, large-scale flood forecasts even in regions with sparse observation networks. It achieves both wide-area coverage and high-resolution projections for critical locations, providing flood depth and inundation extent forecasts several days in advance.
From Weather to Flooding
An End-to-End Forecasting Pipeline
High-Resolution Rainfall Forecasting
Starting from global weather models such as ECMWF, we generate localized high-resolution rainfall data optimized for each region.
Land Surface Modeling
Rainfall data is processed with ILS/MATSIRO to estimate river runoff considering soil infiltration and evapotranspiration.
Large-Scale River Simulation
Using CaMa-Flood developed at the University of Tokyo, we simulate river discharge and water level changes across large basins.
High-Resolution Flood Mapping
Detailed inundation models visualize flood depth at meter-level resolution.
Example of predicted flood depth in the XX river basin.
Left: current time; right: 36-hour forecast. Flood depth (m) is shown using a color scale.
You can compare current and future flood conditions across large river basins in real time. Flood depth and inundation extent are overlaid on maps, enabling local governments and disaster management officials to quickly and intuitively understand the situation.
Use Cases
Proven Forecast Accuracy and Global Track Record
Accuracy Metrics
| Metric | Value | Notes |
|---|---|---|
|
NSE (Nash–Sutcliffe Efficiency) |
0.84 |
Validated at Chavuma, Zambia using ERA5-Land with quantile mapping correction. |
Coverage Areas
- Vietnam (Kon River Basin / Gia Lai Province) — METI Global South Project
- Kyrgyzstan — Climate change adaptation and flood simulation (in collaboration with ArkEdge Space)
- Zambia — JICA Water Resources Master Plan technical demonstration
Research & Technology Partnerships
Joint research with the University of Tokyo Yamazaki Lab (CaMa-Flood / MERIT DEM) and Yoshimura Lab (ILS / MATSIRO Emulator). Recognized as a JAXA partner startup (application in progress). Bias-correction methods implemented using satellite data (TRITON / GSMaP / SWOT / DAHITI).
Onboarding Flow
Hearing & PoC Design
Confirm target region, challenges, and data environment. Design the scope and duration of the proof of concept (PoC).
PoC Execution & Accuracy Verification
Run simulations using actual data. Evaluate forecast accuracy and iterate improvement cycles.
Full Implementation & Ongoing Support
After full system deployment, continue to provide support for operations, accuracy improvements, and training.
First, please tell us about the situation in your region.
The optimal system configuration varies depending on the target river basin, data environment, and operational setup. Initial consultations are free. Please feel free to contact us regarding the feasibility of a PoC.