How to Do Real Environmental Engineering Research at Home: A High School Student’s Guide to Free Tools, Affordable Kits, and Public Databases

Environmental engineering used to mean a university lab, a field truck, and a state-agency badge. Now it means a $30 air sensor on your windowsill, a satellite image on your laptop, and a Python notebook open in the next tab.

This guide is your starting point. It covers three things: the kit you can actually buy and build, the free software that real engineers use, and the public datasets that count as primary data in a science fair report.

Why This Is Possible Now

Three shifts changed everything in the last decade.

First, sensors got cheap. A PM2.5 air-quality node, a conductivity meter, a CO2 logger, a soil-moisture probe: each one costs less than a pair of sneakers. You can build a five-node sensor network for the price of a textbook.

Second, satellite data went open. NASA, ESA, and USGS now stream Landsat, Sentinel, MODIS, SMAP, and TROPOMI imagery to anyone with an internet connection. The same pixels a federal scientist analyzes are the pixels you analyze.

Third, the modeling software that runs city water systems, flood maps, and groundwater plans is free. EPANET, SWMM, MODFLOW, HEC-RAS, and OpenFOAM are all open downloads. They are the actual tools used by consulting firms and government agencies.

Stack those three together and a kitchen counter plus a laptop can monitor a watershed, map an air shed, and simulate a flood.

The Environmental Engineering Home Kit

You will not need everything below. Pick the subset that matches your project shape.

Air and gas sensors

• PM2.5 / PM10 particulate sensors (SDS011, PMS5003): around $20 to $40, plug into an Arduino or ESP32.
• VOC / formaldehyde sensor (PID-style or MQ-series): around $20 to $40 for indoor air projects.
• NO2 and CO electrochemical sensors: around $30 to $60 for traffic-corridor work.
• NDIR CO2 sensor (MH-Z19, SCD30): around $25 to $60 for ventilation and composting projects.
• DHT22 temperature and humidity logger: around $5 to $10 as a baseline for any outdoor experiment.

Water and soil tools

• Conductivity (EC) and pH meters: around $15 to $25 each.
• Turbidity tube or DIY Secchi setup: under $10.
• Aquarium nitrate, nitrite, ammonia, and phosphate test strips: around $15 a pack.
• Mail-in water test kits (ICP-MS for heavy metals, PFAS panels, nitrate panels): around $20 to $150 per sample for results you can defend in a poster.
• Arsenic and fluoride test strips: around $20 a pack for drinking-water projects.

Microcontrollers, loggers, and small hardware

• Arduino Uno or Nano and ESP32 dev boards: around $5 to $15 each. The ESP32 has built-in Wi-Fi and is ideal for sensor nodes.
• Raspberry Pi (3, 4, or Zero 2 W): around $15 to $60. Use this when you need a camera, a real operating system, or on-device machine learning.
• SD card data logger module and real-time clock: around $5 total.
• Piezoelectric buzzers, small pumps, fans, and LEDs for actuator-side projects.

Workshop and wet-lab basics

• 0.001 g jewelry scale for biodegradation, biosorbent, and material mass-balance experiments.
• Kitchen scale (1 g resolution) for evaporation, composting, and yield work.
• Graduated cylinders, syringes, pipettes, and beakers from a basic chemistry set.
• 5-gallon buckets, PVC pipe, storage totes for mini-wetlands, lysimeters, green-roof modules, and infiltration rigs.
• Watering can and protractor for rainfall simulators.
• 3D printer access (school, library, or makerspace) for custom chambers, cyclones, and sensor housings.

Field and imaging tools

• Smartphone with a good camera for colorimetry, photogrammetry, microscopy, and Bortle-class light surveys.
• A second-hand phone or webcam wired to a Raspberry Pi for time-lapse and computer-vision work.
• Hand-held GPS or phone GPS app for georeferencing every sample you collect.

Total realistic cost for a full home kit: about $150 to $400, depending on which sensors your project needs. Many projects come in under $100.

Signature Technique: Low-Cost Sensor Networks

If one technique unlocks more environmental engineering projects than any other, it is building a low-cost sensor network and calibrating it against a reference. Here is the five-step workflow.

1. Pick one pollutant or variable. PM2.5, NO2, CO2, conductivity, soil moisture, or stream stage all work. Write your research question in one sentence first.
2. Build three to five identical nodes. Solder the sensor to an ESP32 or Arduino, add a DHT22 for temperature and humidity, log to an SD card or push over Wi-Fi, and 3D-print a vented enclosure.
3. Co-locate for calibration. Place every node next to a reference monitor for a week. For air quality, the reference can be a PurpleAir station or the AirNow API feed for your nearest regulatory site. For water, use a calibrated handheld or a mail-in test.
4. Fit a correction model. Use scikit-learn to fit a linear, random-forest, or small neural-network correction from your raw readings (plus temperature and humidity) to the reference values. Hold out a test week.
5. Deploy and analyze. Spread the calibrated nodes across your study area (rooms in a house, blocks in a neighborhood, stretches of a creek). Log for at least two to four weeks. Analyze gradients, diurnal cycles, or event responses with pandas and matplotlib.

This shape (cheap hardware plus a calibration layer plus a real spatial or temporal question) is the most reproducible engineering workflow you can run from a bedroom.

The Dry-Lab Side: Free Software You Can Install Today

Hydraulics, hydrology, and flooding

• EPANET: simulates pressurized water-distribution networks. Use it for leak detection, pressure mapping, and sensor placement.
• SWMM (Storm Water Management Model): simulates urban stormwater and combined sewers. Use it for green-infrastructure retrofits.
• HEC-RAS: 1D and 2D river hydraulics and flood inundation. The U.S. Army Corps of Engineers releases it free.
• MODFLOW with FloPy: groundwater flow simulation, scripted in Python.

Geospatial and remote sensing

• QGIS: a full free GIS. Use it for any map, any overlay, any spatial statistic.
• Google Earth Engine: petabytes of satellite imagery with a JavaScript and Python API. Free for students and researchers.
• Open Drone Map: turns a stack of phone or drone photos into a 3D terrain model.
• SNAP: ESA's free toolbox for Sentinel-1 SAR and Sentinel-2 optical processing.

Air, fluid, and process modeling

• OpenFOAM: industrial-grade computational fluid dynamics. Run it on a laptop or a free Colab GPU for street-canyon, HVAC, or cyclone-separator simulations.
• COBRApy: flux-balance analysis for engineered microbial bioremediation pathways.
• openLCA with ecoinvent demo data: full life-cycle assessment for any product or system you can sketch out.

Machine learning and general scientific Python

• Python with NumPy, pandas, scikit-learn, and matplotlib: the backbone of every analysis.
• PyTorch and TensorFlow: deep learning for image classification, LSTM forecasting, and computer vision. Free GPU access through Google Colab.
• BirdNET-Lite: acoustic species classifier that runs on a Raspberry Pi.
• NetLogo and Mesa: agent-based simulation for ecological, recycling, and traffic models.

Running the same software a consulting engineer runs changes how the work feels. You are not doing a "school version" of environmental engineering. You are doing environmental engineering.

Public Databases That Count As Real Data

Air quality

• OpenAQ: aggregated global air-quality measurements with a clean API.
• AirNow API: U.S. EPA regulatory air-quality data, hourly, by station.
• PurpleAir map: dense citizen-science PM2.5 network you can pull historical data from.
• Sentinel-5P TROPOMI: satellite NO2, methane, SO2, and ozone columns.

Water and weather

• USGS National Water Information System: streamflow, groundwater, and water-quality gauge data.
• NOAA Climate Data Online and NOAA Atlas 14: historical weather and design-storm rainfall.
• NASA POWER: long-term solar, temperature, and humidity records for any latitude and longitude.
• NASA SMAP: global soil moisture from space.

Land, vegetation, and remote sensing

• Sentinel-1 (SAR) and Sentinel-2 (optical): free, frequent, high-resolution imagery through Copernicus and Google Earth Engine.
• Landsat: 50-plus-year archive of land-surface imagery from USGS.
• MODIS: daily vegetation, temperature, and fire data.
• VIIRS Day-Night Band: nighttime lights for light-pollution work.

Environmental health, justice, and infrastructure

• EPA EJScreen: demographic and environmental indicators per census block.
• EPA Toxics Release Inventory (TRI): facility-level pollution releases.
• EPA Superfund and Brownfields databases: contaminated-site inventories.
• CDC PLACES and asthma datasets: neighborhood-level health outcomes.
• OpenStreetMap: free, editable map of every road, building, and storm drain.
• i-Tree open data: urban tree benefits and canopy stats.

Re-analyzing public data is a legitimate research path on its own. A clean question, a careful method, and an honest uncertainty estimate beat a new toy every time.

How to Combine Wet and Dry: The Strongest Project Shape

Pattern A: Measure, then model. Build a small physical system (a mini-wetland, a biochar column, a sensor network, an electrocoagulation cell). Collect a clean dataset. Then fit a kinetic model, a breakthrough curve, or a machine-learning predictor in Python. The model lets you extrapolate beyond what you measured.

Pattern B: Model, then measure. Start with a simulation (SWMM for a neighborhood, OpenFOAM for a street canyon, MODFLOW for an aquifer). Identify a sensitive parameter or a hotspot. Then go measure that one thing in the real world to validate the model. The measurement keeps the simulation honest.

Judges respond to this hybrid shape because it mirrors how the field actually works.

Choosing a Phenomenon That Has Not Been Done

Originality is a process, not a guess. Use this three-step novelty check before you commit.

1. Google Scholar. Search your candidate phrasing in quotes. Read the three most recent and three most cited review articles. Note what is settled and what authors call "open questions" or "future work."
2. Society for Science abstracts archive. Search the public ISEF and Regeneron STS finalist abstract databases for keywords from your idea. You are not looking for forbidden topics. You are looking for the closest neighbors so you can position your project differently.
3. PubMed and EPA reports. For bioremediation, water, or health-linked projects, search PubMed. For pollution-control and infrastructure projects, search EPA technical reports and state environmental agency publications.

If you find adjacent work, that is good news. It means your area is alive, fundable, and judge-recognizable. Your job is to find the narrow gap.

A Realistic Timeline

• 1 to 2 weeks: Calibrate one sensor against a reference, or replicate a published kinetic experiment at home with one variable swept.
• 1 to 2 months: A full hybrid project for a regional fair, one wet experiment plus one model, with at least three replicates per condition.
• Full year: An ISEF-track project with a multi-node deployment or a full simulation study, statistical analysis, a clear novel contribution, and a written paper.

If this is your first research project, start with the 1 to 2 week version. Finish it before scaling up.

A Starter Checklist

1. A clean, dedicated workspace with power, ventilation, and a place to leave equipment running for weeks.
2. A free Google account with Google Colab and Google Earth Engine access enabled.
3. A local Python environment (Anaconda or plain Python plus pip) with NumPy, pandas, scikit-learn, matplotlib, and FloPy installed.
4. QGIS installed, plus at least one modeling tool that matches your category (EPANET, SWMM, HEC-RAS, MODFLOW, or OpenFOAM).
5. A bound lab notebook or a dated digital notebook for every measurement, calibration, and code change.
6. A one-sentence research question written at the top of page one.
7. A simple safety plan: gloves, goggles, ventilation, and adult supervision for any pyrolysis, electrochemistry, or hazardous chemical work.

If you have those seven, you are ready to pick a phenomenon.

Where to Go Next

Environmental Engineering at ISEF splits into six subcategories. Pick the one that interests you most. Each has its own MehtA+ project guide that builds on the kit, the software, and the datasets on this page.

• Bioremediation (BIR): Living systems (plants, fungi, bacteria, algae) that clean soil, water, or air.
• Land Reclamation (ENG): Restoring degraded land after mining, fire, salt damage, or urban abandonment.
• Pollution Control (PLL): Engineered systems that remove or prevent air, water, and light pollution at the source.
• Recycling and Waste Management (REC): Turning waste streams into materials, energy, or cleaner inputs.
• Water Resources Management (WAT): Modeling and managing drinking water, stormwater, groundwater, and floods.
• Other (OTH): Cross-cutting projects on environmental justice, sensors, digital twins, biodiversity, and climate resilience.

The kit on a kitchen counter and the laptop on your desk can now do real environmental engineering. Pick a subcategory and start.

Project ideas in this category (68)