Urban Flood Modeling With Satellite Data

By /

Urban Flood Modeling With Satellite Data

ISEF Category: Earth and amp; Environmental Sciences

Ready to Turn This Idea Into a Real Project?

This guide was put together with the help of AI research tools to give you a solid starting point.But a competitive science fair project lives in the details: refining your research question, fine-tuning your variables, analyzing your data, and presenting your findings like a seasoned scientist.

For next steps tailored to your interests, skill level, and timeline, work one-on-one with a MehtA+ mentor. Learn more about MehtA+ Science & Engineering Research Mentorship →

Subcategory: Water Science  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

A flood map can look right and still miss the real flood. That matters when people use it to plan road closures, evacuations, and drainage upgrades. You can test whether a model really matches what satellites saw during past storms.

What Is It?

This project asks a simple question with a hard answer: can a computer model predict where floodwater spread in a real city during a storm? HEC-HMS estimates how rain becomes runoff in a watershed. HEC-RAS then routes that water through channels, streets, and low areas. Think of it like sending water through a digital plumbing system and checking whether the spill pattern matches the real world.

You drive the model with GPM IMERG, which is a satellite rainfall product. Then you compare the modeled flood extent with Sentinel-1 SAR images from past storms. SAR stands for synthetic aperture radar. It can see wet and flooded surfaces even when clouds block normal images. That makes it useful for checking whether your model got the right places wet.

Why This Is a Good Topic

This is a strong science fair topic because you can test a real prediction against an outside data source. You are not just drawing a map, you are checking model accuracy. The project connects to urban flooding, storm response, and drainage planning, which makes the results meaningful. You can also learn data cleaning, GIS, calibration, and validation without needing to invent a brand-new model from scratch.

Research Questions

  • How does using GPM IMERG rainfall instead of local gauge data change flood extent prediction in an urban catchment?
  • What is the effect of changing curve number values on modeled peak runoff and inundation area?
  • Does adding higher-resolution terrain data improve agreement between HEC-RAS flood maps and Sentinel-1 SAR observations?
  • To what extent does model performance change across storms with different rainfall intensity and duration?
  • Which calibration setting gives the best match between modeled and observed flood extents for a specific storm?
  • How does the inclusion of impervious surface data affect runoff volume and low-lying flood zones?

Basic Materials

  • Laptop or desktop computer with enough storage for GIS and model files.
  • HEC-HMS software.
  • HEC-RAS software with 2D modeling capability.
  • GIS software such as QGIS.
  • GPM IMERG precipitation data from NASA.
  • Sentinel-1 SAR flood scenes from NASA or Copernicus data portals.
  • Digital elevation model data from USGS or a local open data portal.
  • Land cover and impervious surface data.
  • Spreadsheet software for tracking parameters and results.
  • External drive or cloud storage for backing up large raster files.

Advanced Materials

  • High-resolution LiDAR-derived elevation data.
  • Stream gauge data for calibration and validation.
  • Local drainage or stormwater network data.
  • Soil data from USDA or a local geospatial database.
  • Building footprint and road network layers.
  • Cloud computing access for batch runs and large raster processing.
  • Python environment for automated parameter sweeps and metric calculation.
  • ImageJ or SNAP for inspecting Sentinel-1 outputs and masks.
  • ArcGIS Pro or advanced QGIS plugins for hydrologic preprocessing.
  • Statistical analysis software for uncertainty testing and sensitivity analysis.

Software & Tools

  • HEC-HMS: Simulates rainfall-runoff processes and estimates how storm water becomes discharge.
  • HEC-RAS: Models channel flow and flood extent in the catchment.
  • QGIS: Prepares elevation, land cover, and output maps for comparison.
  • SNAP: Processes Sentinel-1 SAR scenes and helps extract flood extent.
  • Python: Automates parameter testing, map comparison, and summary statistics.

Experiment Steps

  1. Define one urban catchment and choose a small set of storm events with both rainfall and Sentinel-1 coverage.
  2. Build a data inventory, then check that all rasters, rainfall files, and boundaries use the same coordinate system and time window.
  3. Set the baseline hydrology and hydraulics model structure before you start tuning parameters.
  4. Decide how you will convert both model output and SAR imagery into comparable flood extent maps.
  5. Plan a calibration strategy that changes one parameter group at a time and keeps a separate validation storm untouched.
  6. Choose your accuracy metrics before running the final comparison, so you can report agreement consistently.

Common Pitfalls

  • Using rainfall data and satellite imagery from mismatched storm windows, which makes the model appear worse than it really is.
  • Comparing SAR water masks to raw model water depth without converting both to the same flood extent definition.
  • Ignoring DEM resolution limits, which can hide drainage paths and overstate flooding in flat neighborhoods.
  • Calibrating on the same storm you use for validation, which makes the model look better than it generalizes.
  • Treating cloudy optical flood maps like SAR data, which fails because optical images cannot reliably see through clouds.

What Makes This Competitive

A stronger project does more than make a flood map. It tests uncertainty, compares multiple storms, and explains when the model succeeds or fails. You can also compare different terrain resolutions, land-cover inputs, or calibration methods to see which change matters most. If you add a clean validation design and strong statistics, your project starts to look like real flood research, not just software output.

Project Variations

  • Compare model accuracy across a downtown catchment, a suburban catchment, and a mixed land-use watershed.
  • Test whether LiDAR terrain data improves flood extent prediction more than standard DEM data.
  • Use multiple satellite rainfall products, then compare which one drives the best flood simulation.

Learn More

  • HEC-HMS User Guide: Search the US Army Corps of Engineers HEC documentation site for the official rainfall-runoff model manual.
  • HEC-RAS User Manual: Search the US Army Corps of Engineers HEC documentation site for river and flood modeling guidance.
  • NASA GPM IMERG data: Search NASA Earthdata for Global Precipitation Measurement rainfall products and documentation.
  • Copernicus Sentinel-1 data: Search the Copernicus Data Space or NASA resources for SAR scenes and flood-related guidance.
  • USGS National Map: Search for elevation data, hydrography layers, and terrain resources for US catchments.
  • NOAA National Water Center resources: Search NOAA for flood forecasting and hydrology background material.
Shopping Cart