Research Track 01 · Bluewater Basin ·
Master Scientific Research
Through Real Missions.
Follow the first mission Bluewater Basin expedition to learn scientific thinking, statistics, GIS, Python, R, machine learning, hydrology, hydrogeology, remote sensing and AI through one connected research project from field observations to scientific communication. Other missions would follow.
13
Research Missions
70
Interactive Lessons
100%
Free & Open Source
∞
Endless Possibilities
What you'll practice in Bluewater Basin
Interactive GIS
Explore spatial data with the mapping tools working researchers actually use.
ExploreData Analysis
Analyze real environmental datasets using statistics, from raw readings to insight.
ExploreMachine Learning
Build and train models that predict real environmental outcomes.
ExploreHydrology
Understand water systems, from rainfall to groundwater to ocean interaction.
ExploreOne platform, many fields
Research Atlas isn't just environmental science.
It's a way of teaching research itself, through one continuous, realistic mission simulation instead of disconnected exercises. Bluewater Basin is where that approach lives today. Other research domains are planned to follow the same model, built out with real mission content over time.
Life Sciences
Coming soon
Biology, ecology, and life-science research methods, taught the same mission-based way as Bluewater Basin.
Social & Behavioral Science
Coming soon
Surveys, experiments, and research methods for studying people, communities, and behavior.
Physical Sciences & Engineering
Coming soon
Physics, chemistry, and engineering research methods, modelling, and lab-adjacent data analysis.
How the expedition runs
Every mission follows real scientific workflow.
Ask
Is groundwater near the agricultural zone actually contaminated, or does it look that way by chance?
Observe
Read the terrain, the well logs, and the sampling history before touching a model.
Design
Choose a sampling plan and a method that can actually answer the question.
Analyze
Clean, explore, model, and quantify uncertainty — in that order.
Communicate
Turn the analysis into a figure and a claim someone else can check.
The expedition log · Track 01
Thirteen missions. One watershed.
Every mission is a real milestone in investigating Bluewater Basin — numbered because the order genuinely matters. Mission 7's model depends on the wells you mapped in Mission 1.
- 00
Becoming a Researcher
Meet the Bluewater Basin watershed and the research team. Learn how scientists actually ask questions, form hypotheses, and avoid fooling themselves.
Scientific Thinking · Basin Field Station · ~45 min
- 01
Understanding the Landscape
Read the land like a scientist: elevation, slope, watersheds, and land cover. Build your first map of Bluewater Basin.
GIS & Remote Sensing · Ridge Overlook · ~60 min
- 02
Collecting Environmental Data
Design a sampling plan for rainfall, groundwater, and water quality stations without introducing bias.
Research Design · Rain Gauge Network · ~55 min
- 03
Cleaning Scientific Data
Real sensors fail, drift, and lie. Learn to detect outliers, handle missing values, and document every decision.
Data Science · Data Lab · ~60 min
- 04
Exploring Patterns
Before you model anything, look at it. Distributions, correlations, and the plots that reveal what the basin is doing.
Exploratory Data Analysis · Field Notebook · ~50 min
- 05
Scientific Statistics
Hypothesis testing, confidence intervals, and p-values — grounded in whether Bluewater Basin's water quality is really changing.
Statistics · Water Quality Lab · ~70 min
- 06
Regression & Prediction
Predict streamflow from rainfall. Linear regression, model diagnostics, and why R² can lie to you.
Statistical Modelling · Streamflow Gauge · ~75 min
- 07
Machine Learning
Random forests and cross-validation to predict flood risk across the basin from terrain and rainfall features.
Machine Learning · Flood Risk War Room · ~80 min
- 08
Spatial Analysis & GIS
Kriging, variograms, and spatial autocorrelation to map groundwater contamination between monitoring wells.
Geostatistics · Groundwater Monitoring Grid · ~80 min
- 09
Physics-Informed AI
Blend Darcy's Law with neural networks so predictions never violate the physics of groundwater flow.
Physics-Informed ML · Aquifer Simulation Lab · ~85 min
- 10
Quantifying Uncertainty
Every model is wrong to some degree — learn to say precisely how much, with Bayesian credible intervals and ensembles.
Bayesian Statistics · Uncertainty Chamber · ~75 min
- 11
Scientific Communication
Turn your analysis into a figure and a paragraph that a policymaker, and a peer reviewer, can both trust.
Scientific Writing · Communications Studio · ~55 min
- 12
Publishing Your Research
Package your Bluewater Basin investigation as a reproducible, open, citable piece of research.
Open Science · Basin Field Station — Archive Room · ~50 min
Learn by dragging things
Featured simulations
From the field
What early researchers say
“I'd taken a statistics course twice before and never understood why a p-value mattered. Seeing it decide something real about Bluewater Basin's water quality made it click in one sitting.”
“The fact that Mission 7's random forest uses the exact same wells I mapped in Mission 1 is what made me keep going. It never feels like a new toy example.”
“This is the first place that taught me GIS by making me care whether a well was contaminated, not by making me memorize menu items.”
Before you start