Flood Model - README

Overview

This document provides a systematic guide to developing a comprehensive flood model. The process encapsulated within focuses on leveraging varied datasets, tools, and methodologies to create accurate and actionable flood risk maps.

Workflow:

1. Decision on Natural Event

• Focus: Decide on the specific natural event and region you are analyzing.
• Guidance: Prioritize areas with significant rainfall or water surge events to maximize the model's effectiveness.

2. Define Area of Interest (AOI) & Relevant Dates

• AOI: Define the specific area that experienced the natural event.
• Dates: Specify the date range of the event.

3. Synthetic Aperture Radar (SAR) Data Collection & Processing

• Guidance: Choose the most appropriate band (e.g., VH, VV) based on the region's characteristics and data quality.
• Outcome: Derive a water/flood inundation map from the processed SAR data.

4. Digital Elevation Model (DEM) Processing

• Objective: Obtain elevation data to understand terrain properties, which will guide water flow.

5. Hydrology Analysis in GIS Tools

• Components Analyzed:
  • Slope
  • Water direction & accumulation
  • Distance to streams/channels
• Outcome: Acquire hydrological insights to predict and assess water flow patterns.

6. Landcover Data Collection

• Objective: Understand terrain types, vegetation, and urbanized areas, which impact flood behavior.

7. Collect Rainfall Data

• Guidance: Obtain data from relevant meteorological organizations or databases for the specified AOI and dates.

8. Cloud Optimized Geotiff (COG) Conversion

• Datasets to Convert:
  • DEM
  • Slope
  • Flow accumulation
  • Landcover
  • Precipitation
  • SAR before the event

9. Data Ingestion into Geospatial Platforms

• Objective: Ingest the data into platforms like Google Earth Engine (GEE) for further processing and visualization.

10. Convert Raster Data to Database-Friendly Format

• Objective: Convert processed raster data into tables or other database-friendly formats for analysis and storage.

11. Train/Retrain the Flood Model

• Guidance: Use relevant event data to train or finetune your flood prediction model.

12. Generate Contingency Table

• Objective: Assess the model's performance metrics such as accuracy, precision, recall, etc.

13. Generate Predictions for Other Relevant Areas

• Guidance: Apply the trained model to other regions of interest to understand its generalization capabilities.

14. Generate Optimized Geotiffs for Risk Maps

• Objective: Convert risk maps into geotiff formats optimized for geospatial platforms.

15. Visualize Risk Maps on Geospatial Platforms

• Guidance: Use platforms like GEE or GIS tools for visualization and further analysis.

16. Create Proprietary Flood-risk Maps (If Required)

• Objective: Design and customize flood-risk maps specific to certain platforms or client requirements.

Supplementary Analysis

The flood model can be enhanced by integrating census data and building data from sources like OpenStreetMaps. This ensures a holistic risk assessment, considering population density, infrastructure, and other critical factors.

Best Practices & Notes

• Ensure all datasets have appropriate access rights and permissions.
• Regularly backup data and processed files.
• Validate the model against ground truth or reliable data sources for accuracy.
• Collaborate with domain experts for nuanced insights.

Contributors & Acknowledgements

List all contributors to the project and any key resources or datasets used.

This README serves as a generalized guide. As the process evolves or as unique challenges arise for specific regions, appropriate modifications to the steps and methodologies are encouraged.