Near-Time GIS for Flood Monitoring and Analysis in Niger River Basin, Nigeria

This repository hosts the resources, code, and documentation for the study "Enhancing Disaster Response and Resilience Through Near-Time GIS for Flood Monitoring and Analysis in Niger River Basin, Nigeria", as presented at the ISPRS TC III Mid-term Symposium, 2024.

Overview

Flooding in Nigeria is a critical issue affecting lives, livelihoods, and infrastructure, particularly in the Niger River Basin. This project leverages Google Earth Engine (GEE) for near-real-time flood mapping, impact analysis, and flood resilience modeling. It combines multi-temporal remote sensing datasets with geospatial analysis techniques to quantify flood hazards, assess exposure, and provide actionable insights for disaster response.

Key Features

• Automated Flood Monitoring: Utilizes Sentinel-2 and Sentinel-1 datasets for accurate and timely flood mapping.
• Impact Assessment: Integrates population, land use, and building footprint data to quantify flood impacts.
• Scalable Workflow: Cloud-based processing using GEE for efficient and reproducible geospatial analytics.

Methodology

Data Sources

• Sentinel-2 MSI: Optical imagery for flood detection and land-use classification.
• Sentinel-1 SAR: Radar imagery for cloud-penetrating flood analysis.
• NASA SRTM DEM: Elevation data for terrain analysis.
• Global Surface Water Dataset: Historical water body dynamics.
• Landsat 8: Additional multi-temporal land cover analysis.

Tools and Libraries

• Google Earth Engine: Primary platform for data processing and analysis.
• ArcGIS Pro: For additional preprocessing and visualization.
• Python: Supporting scripts for modeling and data manipulation.

Workflow

1. Data Acquisition: Collect and preprocess satellite imagery and terrain datasets.
2. Flood Mapping: Apply spectral indices (e.g., NDWI) and radar analysis to delineate flood extents.
3. Impact Analysis:
   • Assess changes in land use and land cover.
   • Overlay flood extents with population and building datasets.
4. Visualization: Generate maps and reports using geospatial tools.

Results

• Flood Extent Analysis: A 53% increase in inundation area was observed during the 2022 flood peak.
• Land Use Impacts:
  • Loss of agricultural land and vegetation due to prolonged waterlogging.
  • Conversion of 15,000 hectares to unproductive wetlands.
• Population and Infrastructure:
  • Over 150,000 residents were affected by flooding.
  • 143,000 buildings identified within flood-prone zones.

Maps and Visualizations

Installation and Usage

Prerequisites

• A Google Earth Engine account.
• GIS software like ArcGIS.

Setup

1. Clone this repository:

git clone https://github.com/<your-username>/<your-repo>.git

2. Access the Google Earth Engine scripts provided in the code/ directory.
3. Configure the Area of Interest (AOI): Each script contains a clearly marked CONFIGURATION section at the top. To adapt the analysis to a different location:
   • Replace the AOI / geometry variable with your own ee.Geometry.Polygon coordinates or point it at an Earth Engine asset (e.g., an uploaded Shapefile).
   • Adjust the beforeStart/beforeEnd and afterStart/afterEnd date ranges to match the flood event you are studying.
   • For LULC.js, update the training-data FeatureCollection paths to training samples drawn over your AOI.
   • For Sentinel 1.js, tune the analysis thresholds (diffThreshold, slopeThreshold, connectedPixelThreshold) as needed.

Running the Analysis

• Open the desired script in the GEE Code Editor.
• Paste the script content or import it.
• Update the CONFIGURATION block at the top to match your study area and time period.
• Run the script — results (maps, exports) will be generated automatically.
• Export results (e.g., GeoTIFFs) for further analysis in GIS software.

Applications

This workflow can be adapted for:

• Real-time flood monitoring and early warning systems.
• Climate change adaptation planning.
• Urban and regional flood risk management.

Contributions

Contributions to improve this repository are welcome! Please open an issue or submit a pull request.

Citation

If you use this repository in your research or projects, please cite:

Akintola, M. O. (2024). Enhancing Disaster Response and Resilience Through Near-Time GIS for Flood Monitoring and Analysis in Niger River Basin, Nigeria. ISPRS TC III Mid-term Symposium, Belém, Brazil.

License

This project is licensed GPL-3.0 License