The Flood & Drought Monitoring Application is a Google Earth Engine (GEE)-based geospatial analysis platform designed to monitor, visualize, and assess the impacts of climate-induced hazards including floods and droughts. The application integrates multi-source satellite and geospatial datasets to generate:
- Flood extent maps using Sentinel-1 Synthetic Aperture Radar (SAR)
- Drought condition maps using rainfall anomaly analysis
- Population exposure assessment using WorldPop datasets
- Agricultural impact estimation using ESA WorldCover land-use data
- Interactive hazard visualization through a Google Earth Engine user interface
The system demonstrates how cloud-based remote sensing technologies can support rapid disaster assessment, environmental monitoring, and decision-making.
The primary objectives of this project are:
- Develop an interactive disaster monitoring application using Google Earth Engine
- Detect flood-affected regions using Sentinel-1 SAR imagery
- Identify drought-prone areas using rainfall variability analysis
- Estimate vulnerable population exposure
- Assess potential agricultural losses
- Provide a simple visualization dashboard for hazard interpretation
Geospatial Data Sources
Sentinel-1 SAR CHIRPS Rainfall | | | | ↓ ↓
Flood Detection Drought Detection
WorldPop ESA WorldCover
| |
↓ ↓
Population Exposure Cropland Impact
|
↓
Google Earth Engine Processing
|
↓
Interactive Dashboard
Dataset: COPERNICUS/S1_GRD
Flood detection
- C-band Synthetic Aperture Radar
- Cloud independent observation
- Day/night acquisition capability
- VV polarization analysis
Flood areas are detected based on reduced radar backscatter values:
VV Backscatter < Threshold
↓
Potential Water/Flood Pixels
Dataset:
UCSB-CHG/CHIRPS/PENTAD
Drought monitoring through rainfall anomaly analysis.
CHIRPS provides:
- Satellite-based rainfall estimation
- Long-term precipitation records
- Global coverage
Method:
Mean Annual Rainfall < Threshold
↓
Drought Condition
Dataset:
WorldPop/GP/100m/pop
Population exposure estimation.
The system overlays hazard maps with population distribution:
Hazard Area
Population Density
↓
Exposed Population Estimate
Dataset:
ESA/WorldCover/v100
Agricultural impact assessment.
Cropland classification:
Land Cover Class = 40
↓
Cropland Extraction
↓
Flood/Drought Impact Analysis
- Collect Sentinel-1 SAR imagery
- Filter by:
- Study area
- Date range
- Instrument mode
- Polarization
- Generate median radar composite
- Apply water threshold classification
- Produce flood extent layer
- Collect CHIRPS precipitation data
- Calculate rainfall statistics
- Apply rainfall anomaly threshold
- Generate drought severity map
Hazard maps are combined with:
- Population density
- Land cover classification
to estimate:
- Human exposure
- Agricultural vulnerability
Users can select:
- Floods
- Droughts
- Cropland Loss
- Affected Population
The dashboard provides:
- Hazard maps
- Color-coded layers
- Rainfall time-series charts
- Regional statistics
Flood-And-Drought-Monitoring-App-GEE-Platform/ │ ├── Flood_Drought_Monitoring_App.js │ ├── Project_Code_Structure.js │ ├── README.md │ └── LICENSE
The application generates:
| Output | Description |
|---|---|
| Flood Extent Map | Areas potentially affected by flooding |
| Drought Map | Low rainfall regions |
| Population Exposure | People located in hazard zones |
| Cropland Loss | Agricultural areas under risk |
| Rainfall Chart | Temporal precipitation analysis |
This prototype has several limitations:
- Fixed backscatter threshold
- No terrain correction
- Limited validation with ground observations
- Uses rainfall only
- Does not include vegetation stress or soil moisture
Population exposure does not represent actual casualties or damage.
More detailed models require:
- Flood depth estimation
- Infrastructure data
- Socioeconomic information
Possible improvements include:
Add:
- Random Forest classification
- Deep learning segmentation
- Flood probability prediction
Integrate:
- NDVI anomaly
- Vegetation Condition Index (VCI)
- Soil Moisture Index
- Standardized Precipitation Index (SPI)
Future versions can include:
- Weather API integration
- Automated alerts
- Web-based dashboards
- Mobile notifications
Google Earth Engine has been widely adopted for large-scale environmental monitoring because it provides cloud-based access to satellite archives and scalable geospatial computation.
Relevant applications:
- Flood mapping
- Drought assessment
- Land-cover change detection
- Climate impact analysis
Many studies use Sentinel-1 SAR imagery for flood detection because radar observations are not affected by cloud cover.
Common approaches include:
- Backscatter thresholding
- Change detection
- Machine learning classification
Example research areas:
- Rapid flood mapping after extreme rainfall events
- Near-real-time disaster response systems
CHIRPS rainfall data has been widely used for drought analysis because of its high temporal resolution and long historical coverage.
Common drought indicators:
- Rainfall anomaly
- SPI calculation
- Seasonal precipitation monitoring
Combining hazard layers with population datasets such as WorldPop is a common approach for estimating disaster vulnerability.
Applications include:
- Flood risk assessment
- Humanitarian planning
- Emergency response prioritization
Recent disaster management systems combine:
- Remote sensing
- GIS analysis
- Climate datasets
- Machine learning
to create decision-support tools for governments and humanitarian organizations.
-
Gorelick et al. (2017).
Google Earth Engine: Planetary-scale geospatial analysis for everyone.
Remote Sensing of Environment. -
Torres et al. (2017).
Sentinel-1 SAR mission and applications. -
Funk et al. (2015).
The Climate Hazards Infrared Precipitation with Stations (CHIRPS). -
Tatem et al.
WorldPop: High-resolution population datasets for development and disaster applications. -
Zanaga et al. (2021).
ESA WorldCover 10 m land cover dataset.
Contributions are welcome.
You can contribute by:
- Improving hazard detection algorithms
- Adding new datasets
- Improving visualization
- Adding validation methods
Steps:
- Fork the repository
- Create a feature branch
- Commit changes
- Submit a pull request
Sohel Rana
Digital Currency Investor & Technical Lead
https://github.com/SohelRana-aiub-Pro/Flood-And-Drought-Monitoring-App-GEE-Platform/
Related Resources; https://github.com/ianpdavies/cloudy_flood_prediction