Beaver-Dam Analogs and Stream Baseflow

ISEF Category: Earth and Environmental Sciences

Subcategory: Water Science  ·  Difficulty: Advanced  ·  Setup: Home Setup  ·  Time: Full Year

The Hook

A small dam can change a stream more than you might expect. Beaver-dam analogs are built to slow water, spread it out, and keep streams wetter for longer. You can test whether that actually raises baseflow using satellite data and stream gages. That means your project can move from a local restoration idea to real evidence.

What Is It?

Beaver-dam analogs, or BDAs, are human-built structures that copy some effects of beaver dams. They slow water, trap sediment, and can raise the water table near a stream. Think of them like tiny speed bumps for runoff. Instead of water racing downstream, some of it lingers, so the stream may keep flowing during dry periods.

Baseflow is the part of streamflow that comes from groundwater, not direct rain or snowmelt. If BDAs help water soak into the ground, baseflow may rise later on. Sentinel-2 gives you repeat satellite images, and the NDWI, or Normalized Difference Water Index, helps estimate how wet or water-covered an area looks. USGS gages give you streamflow records, so you can compare changes before and after BDA installation.

Why This Is a Good Topic

This is a strong science fair topic because you can test a real environmental intervention with public data. You do not need a wet lab, but you do need to think like an ecologist and data analyst. The question connects to drought resilience, stream restoration, and water storage. You can learn how to define controls, build a before-and-after comparison, and test whether a change is likely linked to the BDAs rather than weather alone.

Research Questions

• How does BDA installation change baseflow at treated stream reaches compared with similar untreated reaches?
• What is the effect of BDA installation on NDWI patterns near the stream corridor over time?
• Does the size of the baseflow change differ between wet seasons and dry seasons after BDA installation?
• To what extent do precipitation anomalies explain the same streamflow changes that BDAs appear to cause?
• Which distance from the stream centerline shows the strongest NDWI response after BDA installation?
• How does the baseflow response vary between streams with different watershed sizes or slopes?

Basic Materials

• Laptop with internet access.
• USGS National Water Information System gage records.
• Sentinel-2 imagery from Copernicus data access portals or Google Earth Engine.
• GIS software such as QGIS.
• Spreadsheet software such as Google Sheets or Excel.
• List of BDA sites and matched control sites from public project reports or agency documents.
• USGS streamflow and precipitation data for the study period.

Advanced Materials

• High-performance laptop or cloud computing access for time-series analysis.
• Google Earth Engine account for large-scale Sentinel-2 processing.
• QGIS or ArcGIS Pro for spatial analysis.
• Python with pandas, geopandas, rasterio, statsmodels, and matplotlib.
• Digital elevation model data from USGS EarthExplorer or NOAA sources.
• National Land Cover Database layers for watershed context.
• Stream temperature or groundwater datasets, if available from agency partners.

Software & Tools

• QGIS: Maps BDA sites, watershed boundaries, and stream buffers for spatial comparison.
• Google Earth Engine: Pulls Sentinel-2 scenes and computes NDWI time series over long periods.
• Python: Cleans gage records, joins weather data, and runs the difference-in-differences model.
• R: Fits regression models and checks assumptions for causal comparison.
• ImageJ: Measures water or vegetation signal in small image subsets if you build a manual backup workflow.

Experiment Steps

1. Define one treated stream reach with BDAs and one or more matched control reaches with similar watershed traits.
2. Collect a clean before-and-after timeline for satellite imagery, streamflow records, and precipitation records.
3. Build an analysis window that isolates the stream corridor and separates it from nearby land cover.
4. Decide how you will turn NDWI and gage data into baseflow indicators, then plan a standard comparison metric.
5. Set up the difference-in-differences model so you can compare treatment and control trends before and after installation.
6. Plan sensitivity checks that test whether your result holds across seasons, buffer sizes, and site matches.

Common Pitfalls

• Comparing a treated stream to a control stream with a very different watershed, which makes the causal test weak.
• Using satellite scenes with clouds or snow, which can distort NDWI and hide the water signal.
• Treating total streamflow as baseflow without separating storm response, which can blur the effect you want to measure.
• Ignoring precipitation timing, which can make a wet year look like a BDA effect.
• Picking only one site and one short time window, which leaves you with too little evidence to separate trend from treatment.

What Makes This Competitive

A stronger version of this project uses careful matching, not just a before-and-after graph. You can compare multiple control reaches, test whether the parallel trends assumption holds, and report effect sizes with confidence intervals. A competitive entry also checks whether the result changes with season, drought level, or watershed type. That kind of analysis shows you understand both the ecology and the statistics.

Project Variations

• Use only one watershed and compare multiple BDA clusters within it to reduce site-to-site climate differences.
• Swap NDWI for another water-related spectral index and test whether the signal changes for narrow riparian zones.
• Focus on drought periods only and test whether BDAs matter more when baseflow is already low.

Learn More

• USGS National Water Information System: Search for stream gage records, discharge data, and site metadata for your study streams.
• USGS Water Science School: Learn the basics of streamflow, baseflow, and watershed processes from a free government source.
• NASA Earthdata: Find tutorials and documentation for Sentinel-2 style remote sensing workflows and spectral indices.
• Copernicus Sentinel-2 User Guide: Read about the bands and image properties needed for NDWI calculations.
• NIH PubMed: Search for review articles on beaver-dam analogs, stream restoration, and hydrologic response.
• QGIS Documentation: Learn free GIS mapping, buffering, and raster handling for your site comparison.