Advanced Surface Water Quality Modeling

Why this course

Water quality models drive regulatory decisions: how much pollution a river can take, and who has to cut back. This course teaches students to build those models from the ground up, from a single mass-balance box to a calibrated SWAT or HSPF watershed. It also teaches them to defend those models, which means getting the right answer for the right reason. The final module sets process-based models side by side with AI, component-based modeling, and data assimilation, so students can judge where each belongs.

Labs follow one running case study: the “Verde-Agua” watershed, a basin shifting from intensive agriculture to suburban development, with eutrophication in its terminal reservoir.

Who it’s for

Graduate students in water resources, environmental engineering, and related fields.

Prerequisites

• CEE 440/545 Hydrology or an equivalent introductory hydrology course
• A background in quantitative analysis
• Basic programming (Python or Excel scripting) and familiarity with mass balance

What you’ll be able to do

1. Construct and evaluate lumped watershed models for water, sediment, and nutrient mass balances.
2. Implement distributed watershed models such as SWAT and HSPF.
3. Calibrate model parameters and run sensitivity analyses with parameter estimation tools (PEST).
4. Critically analyze research literature on advanced modeling techniques.
5. Explain the role of AI, component-based models, and data assimilation in watershed modeling.
6. Design and defend a modeling approach for a real watershed problem.

Course plan

1. Weeks 1–4

   Fundamentals of watershed modeling and nonpoint pollution

   • Watershed processes; point vs. nonpoint sources
   • Mass balance and lumped models (CSTR, first-order kinetics)
   • Sediment and nutrient transport, sources, and fate
   • Building and stress-testing a lumped model in Python or Excel
2. Weeks 5–10

   Process-based modeling

   • Watershed delineation and input data
   • SWAT and HSPF structure and setup
   • Simulating flow, sediment, and nutrients; management scenarios
   • Calibration and sensitivity analysis with PEST
   • AI/ML in watershed modeling: SWAT vs. machine learning
3. Weeks 11–13

   Advanced modeling concepts

   • Component-based modeling with Landlab
   • Data assimilation (ensemble Kalman and particle filters)
   • Critical paper review (seminar format)
4. Weeks 14–15

   Class project: a TMDL for an impaired waterbody

   • The TMDL framework: margin of safety and load allocations
   • Project studio and presentations

The project

Each student develops a Total Maximum Daily Load (TMDL) for an impaired waterbody of their choosing. The work includes justifying the modeling approach and tools, facing the messiness of real, missing data, and defending the plan in a final report and presentation.

Readings

• Chapra, Surface Water Quality Modeling (suggested)
• Novotny, Water Quality: Diffuse Pollution and Watershed Management, 2nd ed. (suggested)
• Model documentation for SWAT, HSPF, SWMM, and Landlab; selected journal articles