Finding Hidden Groundwater Depletion Hotspots

ISEF Category: Earth and Environmental Sciences

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

The Hook

Some places lose groundwater long before anyone sees a dry well. Satellites can catch that loss from space by tracking tiny changes in Earth’s gravity. When you pair that signal with land model data, you can estimate where water is disappearing underground. That makes this topic a strong choice if you want to study a real water problem with real-world stakes.

What Is It?

GRACE-FO is a pair of NASA satellites that measure changes in Earth’s gravity field. Gravity shifts when water moves around. If a region stores less water underground, in soil, or as snow, the satellites can pick up that change as a mass signal.

That signal by itself does not tell you where the water went. That is where GLDAS helps. GLDAS is a land surface model that estimates snow water, soil moisture, and other surface water stores. If you subtract those surface components from total water storage, the leftover signal can point toward groundwater change. Think of it like checking a full bank account, then subtracting your known expenses. What remains is the part you did not directly see.

For your project, you would compare these datasets over time and look for places where groundwater appears to drop faster than nearby land or climate patterns would suggest. You can then map hotspots in the Midwest or Great Plains and test whether they line up with drought, irrigation, or land use patterns.

Why This Is a Good Topic

This topic works well because the data already exist, the question is real, and the analysis can still feel original. You can test a clear pattern, compare regions, and look for hidden groundwater loss that matters for farming and water planning. You will learn remote sensing, anomaly analysis, and how scientists combine datasets to estimate something they cannot measure directly from the ground.

Research Questions

• How does groundwater storage anomaly change across selected Midwest or Great Plains counties over time?
• What is the effect of subtracting snow and soil moisture components from total terrestrial water storage on hotspot detection?
• Does groundwater depletion align more closely with drought indices or with irrigated land extent?
• To what extent do different baseline periods change the locations of detected depletion hotspots?
• Which counties or grid cells show the largest persistent negative groundwater anomalies?
• How does seasonal variation affect the contrast between groundwater loss and surface water storage change?

Basic Materials

• Computer with internet access.
• Spreadsheet software such as Excel or Google Sheets.
• Free GIS software such as QGIS.
• NASA Earthdata account for downloading GRACE-FO products.
• GRACE-FO terrestrial water storage anomaly data.
• GLDAS land surface model data.
• County or state boundary shapefiles.
• USDA crop or irrigation data, if you compare land use patterns.

Advanced Materials

• Computer with internet access.
• Python with pandas, xarray, numpy, scipy, and matplotlib.
• Jupyter Notebook or similar notebook environment.
• QGIS or ArcGIS Pro if available.
• NASA GRACE-FO mascon products.
• GLDAS Noah or similar land surface model outputs.
• Drought index datasets such as SPEI or US Drought Monitor archives.
• USGS groundwater well observations for validation.
• High-resolution land cover or irrigation datasets from USDA or other public sources.

Software & Tools

• QGIS: Maps groundwater anomaly patterns and helps you compare them with county or basin boundaries.
• Python: Automates data cleaning, anomaly calculations, and hotspot screening.
• NASA Earthdata Search: Finds and downloads GRACE-FO and related Earth observation products.
• Google Earth Engine: Supports large-scale spatial comparison if you want to explore land cover or drought layers.
• ImageJ: Not needed here, so skip it unless you build a separate visual analysis project.

Experiment Steps

1. Define the groundwater question you want to answer, then pick a region, time span, and spatial scale you can defend.
2. Choose one GRACE-FO product and one GLDAS product, then confirm that their time windows and units can be compared.
3. Decide how you will separate total water storage into snow, soil moisture, and groundwater-related residuals.
4. Build a simple anomaly baseline so you can compare wet, dry, and normal periods on the same scale.
5. Plan your hotspot rule, such as persistent negative anomalies or clusters that stay low across multiple months.
6. Design validation checks with outside data, such as drought maps, irrigation patterns, or groundwater well records.

Common Pitfalls

• Mixing GRACE-FO and GLDAS products with different spatial resolutions, which can create false hotspot boundaries.
• Treating the groundwater residual as a direct measurement, when it is actually an estimate after subtraction.
• Ignoring seasonal snow and soil moisture swings, which can hide the true groundwater signal.
• Using a short time window, which makes one drought year look like a long-term depletion trend.
• Comparing counties or grid cells without normalizing for area, which can make large regions look worse just because they are larger.

What Makes This Competitive

A stronger project goes beyond a simple map. You can test whether your hotspot results stay stable across baseline choices, model versions, and seasons. You can also compare the satellite-based estimate with independent groundwater or drought data. That kind of careful validation, plus a clear explanation of uncertainty, makes the work feel much closer to real hydrology research.

Project Variations

• Focus on one state and compare irrigated versus non-irrigated counties to see whether depletion hotspots cluster near agriculture.
• Use drought index data as a second layer and test whether groundwater anomalies lag behind meteorological drought.
• Compare two land surface models or two baseline periods to see how sensitive hotspot detection is to the decomposition method.

Learn More

• NASA Earthdata Search: Download GRACE-FO and GLDAS products, and search the product pages for documentation and user guides.
• NASA GRACE Mission pages: Read background material on how gravity satellites measure changes in water storage on NASA’s GRACE and GRACE-FO pages.
• PubMed: Search for review articles on GRACE-FO groundwater depletion and land water storage anomalies.
• USGS Water Data for the Nation: Find groundwater well records and hydrology context for validation.
• NOAA National Centers for Environmental Information: Look up drought indices, climate normals, and background climate data.
• MIT OpenCourseWare Hydrology or Remote Sensing courses: Review free lecture notes for data analysis and Earth observation basics.