Case Studies

Lecture

James Doss-Gollin

Monday, April 13, 2026

Draft: This content is under development.

The Problem with Perfect Plans

In 1953, catastrophic flooding killed over 1,800 people in the Netherlands.

The Dutch response: the Delta Works — a massive system of dams, sluices, and storm surge barriers.

Designed to last 200 years.

Target: 1:02. Open with the hook — no framing yet, just the story. The Delta Works are an engineering marvel. But they were designed under a paradigm of “find the optimal solution and build it.” What happens when the assumptions behind that solution change? Sea-level rise projections in 1953 looked very different from today’s. Let the students sit with the tension: you can’t just optimize once for infrastructure that lasts centuries.

You’ve Spent 13 Weeks Building a Toolkit

Today we look at how someone actually used it.

A real case study of climate adaptation planning in the Netherlands — one that’s become a model for coastal cities around the world.

Then we turn the lens on your own projects.

Target: 1:05. This is the hook and framing slide. Keep it brief. The point: students have been learning methods in isolation. Today is about synthesis. Emphasize the celebration: they have actually built a serious toolkit. That’s worth acknowledging.

The Adaptive Pathways Approach

Today

1. The Adaptive Pathways Approach

2. Methods in the Wild

3. Reading a Case Study

4. Your Final Projects

5. Looking Ahead

Target: 1:06. Section break. Transition to the Haasnoot case study.

The Core Challenge

The Netherlands faces deep uncertainty about:

• How much sea level will rise (and when)
• How socioeconomic conditions will change
• Whether existing infrastructure will be adequate in 50 or 100 years

Large infrastructure investments have very long lifetimes and high lock-in.

Target: 1:09. This motivates why standard optimization fails here. Lock-in is key: once you build a dike to a certain height, you can’t easily unbuild it. The decision is irreversible (or very costly to reverse), and the uncertainty is deep. Connect to Week 8: this is precisely the setting where robustness frameworks matter most.

The Key Insight

There is no single “optimal” plan under deep uncertainty.

Instead: design pathways — sequences of actions that can be adjusted as we learn.

A pathway is a sequence of decisions where transitions are triggered by observable conditions.

Target: 1:12. This is the conceptual core of the lecture. Pathway = sequence of actions + adaptive triggers. The trigger might be: “when observed sea level rise exceeds X cm” or “when flood frequency exceeds Y per decade.” The key word is “observable” — the trigger has to be something you can actually measure. Draw this on the board if helpful: a timeline with decision points branching into multiple futures.

A Pathway Map

Note

Pathway Map — a diagram showing possible sequences of actions across time, with transitions triggered by observed conditions.

• Each path represents a sequence of policy decisions
• Trigger conditions mark when to switch from one action to another
• Multiple paths accommodate multiple possible futures
• No single path is “the plan” — the map is the plan

Target: 1:15. The pathway map is the central deliverable in Haasnoot et al. Walk through what it looks like: a horizontal timeline, branching action sequences, trigger thresholds marked. Key conceptual shift: the output of the analysis is not a recommendation but a structured menu of options with conditions for choosing among them. This is very different from “find the optimal policy.”

What Makes Pathways Different

Traditional Approach	Adaptive Pathways
Optimize for one scenario	Perform under many scenarios
Static plan	Triggers for adaptation
Single recommendation	Menu of options
Uncertainty as risk	Uncertainty as driver of design

Target: 1:17. A quick contrast to anchor the idea. Don’t dwell — students should see this as a synthesis of ideas they already know. The bottom row is the most important: uncertainty isn’t something to minimize or quantify — it’s the thing that shapes how you design the system.

Methods in the Wild

Today

1. The Adaptive Pathways Approach

2. Methods in the Wild

3. Reading a Case Study

4. Your Final Projects

5. Looking Ahead

Target: 1:18. Section 2: show how this case study uses each Module 3 tool. The goal is for students to see their course content as an integrated pipeline, not a list of separate methods.

Exploratory Modeling → Identify the Landscape

Run the system model across a large ensemble of futures:

• Vary: sea-level rise trajectories, precipitation, demand, socioeconomic growth
• Not: calibrate to a single best estimate
• Goal: map out where current strategies succeed and where they fail

Target: 1:21. This is Week 10 content: exploratory modeling. The emphasis is on mapping the performance landscape, not finding the optimal solution. Key framing: you’re not asking “what’s the best plan?” — you’re asking “under what conditions does any given plan fail?”

Scenario Discovery → Find the Triggers

Use PRIM or similar to find:

“Under what combinations of future conditions does the current strategy become inadequate?”

Those failure conditions become the trigger thresholds in the pathway map.

Target: 1:24. This is Week 10 content: PRIM / scenario discovery. The connection is direct: scenario discovery finds the conditions that define decision points. Example: “Strategy A fails when SLR > 40 cm AND population growth > 1.5%/year.” That box-shaped region in scenario space becomes a trigger: “switch to Strategy B when you observe these conditions.” This is a beautiful integration — scenario discovery literally generates the adaptive triggers.

Sequential Decisions → Design the Pathways

Once you know the triggers, design the sequences:

• What actions are available at each decision point?
• What are the path dependencies? (some options foreclose others)
• Which sequences are robust across many futures?

This is Week 11 content applied at the portfolio level, not just the individual decision level.

Target: 1:27. Week 11: sequential decision making. Path dependencies are important: if you build a very tall dike now, you may foreclose the option to later convert to managed retreat. The pathway map visualizes these dependencies explicitly. “Portfolio level”: instead of optimizing one sequence, you design a set of sequences that together cover many futures.

Multi-Objective → Evaluate the Pathways

Pathways are evaluated on multiple objectives simultaneously:

• Cost: upfront and lifecycle
• Safety: flood risk reduction across scenarios
• Flexibility: how many options remain open?
• Co-benefits: ecosystem services, recreation, equity

No single pathway dominates on all objectives → trade-off visualization (Week 12).

Target: 1:30. Week 12: multi-objective optimization. The pathway map doesn’t produce a single recommendation — it produces a Pareto front of pathways. Decision-makers can then navigate trade-offs explicitly: “do we want the cheapest pathway, or the one that keeps the most options open?” Flexibility is worth calling out: this is a genuine objective, not just a constraint. Some pathways are more reversible than others, and that has real value under deep uncertainty.

The Integrated Analysis Pipeline

Exploratory Modeling
        ↓
  large ensemble of futures

Scenario Discovery
        ↓
  trigger conditions (when to adapt)

Sequential Decision Design
        ↓
  pathway sequences (what to do)

Multi-Objective Evaluation
        ↓
  pathway trade-offs (which to recommend)

Target: 1:32. This is the key synthesis slide of the lecture. The pipeline shows how the four Module 3 methods connect into a coherent workflow. Take a moment here: ask students if they can trace their own project through this pipeline. Not every project will use all four stages, but most will touch at least three.

Reading a Case Study

Today

1. The Adaptive Pathways Approach

2. Methods in the Wild

3. Reading a Case Study

4. Your Final Projects

5. Looking Ahead

Target: 1:33. Section 3: prepare students for Wednesday’s discussion with Dongwook. Give them analytical vocabulary to bring to the paper.

Five Questions for Any Case Study

1. Decision context: Who decides? What’s at stake? What are the constraints?
2. System model: What’s captured? What’s left out?
3. Uncertainty characterization: Deep or well-characterized? How sampled?
4. Decision framework: Static or adaptive? Single or multi-objective?
5. Limits: What does this analysis NOT answer?

Target: 1:36. These five questions give students a scaffold for Wednesday’s discussion. They should arrive having thought through each of these for Haasnoot et al. Question 5 is often the most revealing: good papers are honest about what they can’t answer. For Haasnoot: what’s NOT captured? Implementation costs? Political feasibility? Equity across communities? Encourage students to push on the limits — that’s where the interesting science lives.

Discussion Prep: Haasnoot et al. (2012)

For Wednesday: Dongwook leads discussion on Haasnoot et al. (2012)

Come ready to address:

• What is the specific decision the pathways are designed for?
• What triggers are proposed, and how were they derived?
• What trade-offs does the paper surface — and which does it leave aside?
• How would you apply this approach somewhere other than the Netherlands?

Target: 1:38. Specific prep questions for Wednesday. Dongwook will lead — your job is to set the stage so students come prepared. The last question (transferability) is particularly interesting: the Netherlands has specific institutional capacity, data quality, and political will. Does the method transfer to, say, a small island developing state? Don’t answer these here — just plant the questions.

Your Final Projects

Today

1. The Adaptive Pathways Approach

2. Methods in the Wild

3. Reading a Case Study

4. Your Final Projects

5. Looking Ahead

Target: 1:39. Section 4: connect everything to the final projects. This should feel energizing, not stressful. Students have done the work — now help them see it.

You’ve Already Done the Analysis

Your project follows the same logic as Haasnoot et al.:

• Memo 1: Characterized uncertainties (X) and levers (L); built your model (R) and chose metrics (M)
• Memo 2: Evaluated evidence and robustness across scenarios
• Now: Synthesize into a coherent recommendation (or structured trade-offs)

The pipeline you’ve built is the analysis. The presentation is how you communicate it.

Target: 1:41. Reassurance slide: students have done substantial work. The framing here is important: their memo work wasn’t just homework — it was the actual analytical pipeline. The final presentation is not a new task; it’s communicating what they’ve already done.

The Executive Briefing Challenge

A decision-maker doesn’t want to hear about your methodology.

They want to know:

• What’s at stake — why should they care?
• What you found — what does the evidence say?
• What to do — your recommendation, even if hedged
• What could change your mind — the key uncertainties

Target: 1:43. This is the transition from “technical analysis” to “communication.” The executive briefing is a genre with its own conventions. The biggest mistake students make: leading with methods. No one cares. Lead with the decision. Lead with what’s at stake. Then bring in the evidence. “What could change your mind” is actually a sign of analytical sophistication, not weakness.

Slide Structure Suggestion

Slide	Content
1	What’s at stake — the decision, the stakes
2	The key uncertainty — what you don’t know and why it matters
3–5	What you found — your main results
6–7	Recommendation — what to do, with conditions and limits

15 minutes goes fast. Cut ruthlessly. Practice out loud.

Target: 1:45. Practical slide structure advice. 7 slides for 15 minutes is about right — roughly 2 minutes per slide. “Cut ruthlessly” is the key message. Most students try to show everything they did. The presentation is not a summary of your work; it’s a persuasive argument for a recommendation. Practice out loud: this is not optional. Timing matters, and you can’t know your timing without actually speaking.

One Reflection Prompt

Think about your project for 60 seconds:

What is the single most important figure or finding you want your audience to remember after they leave the room?

That figure should be on slide 3 or 4. Everything else supports it.

Target: 1:47. Give students 60 seconds of actual silence to think. This is a design exercise: what is the load-bearing finding? If they can’t answer in 60 seconds, their narrative isn’t clear yet — and that’s a signal to work on. After the pause, cold call 2–3 students to share their one key finding. This doubles as a preview of what presentations will cover next week.

Looking Ahead

Today

1. The Adaptive Pathways Approach

2. Methods in the Wild

3. Reading a Case Study

4. Your Final Projects

5. Looking Ahead

Target: 1:48. Closing section — brief.

This Week and Next

• Wednesday: Dongwook leads discussion of Haasnoot et al. (2012) — Come with your five questions answered
• Friday, Apr 17: Slides due — Treat this deadline seriously; practice requires finished slides
• Week 14 (Apr 20–24): Executive Briefings — Treat it like a real client presentation

This is the end of the formal curriculum.

Everything from here is yours to use.

Target: 1:50. Close with genuine warmth — this is the last lecture. The final line (“everything from here is yours to use”) is the send-off. The toolkit they’ve built — Monte Carlo, robustness, scenario discovery, multi-objective trade-offs — is genuinely useful in the world. That’s worth saying out loud. Optional: brief acknowledgment that you’ve enjoyed the semester, invite questions in office hours, etc.

References

Haasnoot, M., Middelkoop, H., Offermans, A., Beek, E. van, & van Deursen, W. P. A. (2012). Exploring pathways for sustainable water management in river deltas in a changing environment. Climatic Change, 115(3), 795–819. https://doi.org/10.1007/s10584-012-0444-2