From Watts Up With That?

By Anthony Watts

The University of Copenhagen press release claiming that “over half of global coastal settlements are retreating inland due to intensifying climate risks” sounds like the sort of headline crafted to feed an increasingly popular narrative: coasts are becoming uninhabitable battlegrounds of climate doom, and people are fleeing for safety.

Like many such claims, it rests on a selective look at data, a sweeping interpretation of human motives, and an assumption that climate hazards dominate decision-making for developers and residents. In reality, economics—not climate fear—drives coastal development, and has for centuries. The study’s reliance on nighttime light data as a proxy for population movement adds yet another layer of uncertainty—particularly in developing regions where electrification is uneven.

The study itself if paywalled at Nature Communications, and I’m not going to throw good money at bad science. Fortunately, the press release itself provides enough clues to see the cracks in the narrative.

(All quotations below come directly from the press release.)

1. Developers Don’t Use “Climate Risk Models” to Choose Where to Build — They Follow Money, Not Models

The press release repeatedly frames changes in settlement patterns as “responses to intensifying climate risks.” But nowhere does it present evidence that actual developers, property investors, or residents are citing climate risk as their primary rationale for new construction.

That is because, almost without exception:

Developers build where buyers want to buy. Buyers choose based on economic opportunity, amenities, and cost—not abstract climate projections.

A few points worth emphasizing:

• Coastal real estate in the U.S., Europe, and Asia continues to command premium prices, rising faster than inland regions in many markets.
• Major insurers still willingly underwrite billions in coastal property—usually at higher premiums, but still entirely viable.
• Municipalities encourage coastal development because it boosts tax revenue.
• New flood defenses, sea walls, and drainage upgrades are being built all over the world—not as signs of retreat, but evidence of investment in staying put.

In short: if a region truly believed climate risks were rendering their coasts unlivable, the cost of coastal land would collapse. It hasn’t.

The study authors assume causation: “retreat” must mean fear of climate hazards. But unless they interviewed developers, tracked planning documents, or reviewed permit applications (they did none of these), the causal claim is pure inference, not evidence.

2. The “Nightlights = Retreat” Assumption Is a Giant Leap — Especially Where Electrification Is Uneven

The entire study hinges on a questionable premise: that satellite-detected nighttime light intensity is a reliable indicator of population movement toward or away from the coast.

Even the authors admit the weaknesses:

“Night-time light data may not tell the whole story, as economic activity and settlements extent are not necessarily linked to luminous activity in regions with limited electrification.”

This applies not just to Africa but to large parts of Asia and South America as well. If electrification improves inland faster than along the coast—because of grid expansion, economic development, industrial siting, or government investment—the “retreat” the study claims to detect might simply be the movement of electricity, not people.

In wealthy regions, the situation is reversed: coastal areas often have higher light saturation due to tourism, port operations, and commercial zoning. A modest shift in where industrial lighting is concentrated could dramatically skew the appearance of “retreat” or “advance.”

Nightlights are a useful tool, but only when paired with direct measurements of actual population counts—census data, property tax rolls, verified construction records.
The press release makes clear the study did not do this.

Without validating the light data against real-world settlement patterns, the resulting conclusions are, at best, speculative.

3. Coastal Population Trends Historically Show Growth, Not Retreat — Especially Relative to Inland Growth

To understand whether coastal communities are truly “retreating,” a simple question must be asked:

Are coastal populations shrinking relative to national population growth? Or are they growing, but at a slightly slower rate than in some inland regions?

There is a major difference between:

• People abandoning the coast, and
• Inland development simply growing faster (e.g., suburban sprawl, new industrial zones, tech corridors).

Historically, coastlines have grown faster than inland regions for centuries:

• In the U.S., coastal county populations have increased steadily since at least 1900.
• In Europe, major coastal cities—from Copenhagen to Rotterdam to Barcelona—continue expanding.
• Southeast Asia, despite frequent typhoons, has seen explosive growth in coastal megacities.

The press release itself hints at the real driver:

“For centuries, coastlines have attracted dense human settlement and economic activity.”

That hasn’t changed. Economic opportunities still cluster along the sea.

Before claiming a “retreat,” the authors should demonstrate:

1. That actual population numbers in coastal zones have decreased, not just shifted slightly inland.
2. That growth inland is not simply outpacing coastal growth because of cheaper land or new infrastructure, not climate fear.

The press release presents no such evidence.

4. Middle-Income Regions “Retreating” May Simply Be Developing New Land — Not Fleeing Climate Risks

A key claim:

“Middle-income countries… possess enough institutional capacity and financial resources to support relocation.”

Or—more plausibly—they possess the resources to expand infrastructure inland:

• New highways
• New industrial parks
• New housing blocks
• New suburban developments

When you pave a new ring road 5–15 km inland, suddenly the lights there brighten, and the satellite interprets this as “retreat,” even if the coastal population remains stable or grows.

Economic development ≠ climate retreat.

Urban sprawl ≠ climate retreat.

Electrification ≠ climate retreat.

Unless the authors can show that coastal property is being abandoned—and not replaced by new development—the narrative of retreat is unsupported.

5. The Press Release Claims Retreat Is Driven by “Vulnerability” — But Vulnerability Is Itself a Social, Not Climatic, Metric

From the authors:

“Coastal retreat mostly happens in places where communities don’t have the means to protect themselves… not necessarily because of more hazards.”

This is a revealing admission.

If retreat occurs not because hazards are increasing, but because:

• infrastructure is lacking,
• governments are limited in capacity,
• economies are weak,

…then the phenomenon is, by their own acknowledgment, a development issue, not a climate one.

In fact, this is exactly what historical settlement patterns show:

• Wealthier nations invest in sea walls, surge barriers, and coastal defenses.
• Poorer nations lack these resources and develop unevenly.
• Some regions urbanize inland because it is simply cheaper and less congested.

None of this indicates an accelerating climate-driven retreat.

6. The Claim That “56% of Regions Retreated” Requires Context the Study Doesn’t Provide

The press release claims:

“56% of coastal regions have retreated from the coast from 1992 to 2019.”

This is statistically meaningless without answers to the following:

1. What is the magnitude of the retreat?

A shift of 200 meters inland shows up in satellite light data, but is irrelevant to real-world settlement patterns.

2. What is the baseline distance?

If a region’s bright zone expanded inland due to suburbanization, but coastal density remained stable, that is not retreat.

3. Were coastal lights dimming, or inland lights brightening?

Uneven electrification alone can explain large portions of the “retreat.”

4. Are coastal populations declining?

The press release provides no evidence of shrinking populations—just shifts in illumination.

Without these details, the 56% figure tells us nothing about actual human movement.

7. Copenhagen “Moving Toward the Coast” Undermines the Claim That Hazards Are Increasing

The press release unintentionally highlights a contradiction:

“Denmark… is amongst the minority of regions moving closer to the coast these past decades.”

Denmark—a country with allegedly “accelerating sea-level rise”—is building closer to the water. So are the Netherlands, Singapore, Dubai, Tokyo, Miami, and New York.

When wealthier nations with access to the best data refuse to retreat from the coast, it reveals something important:

Their actions contradict the narrative of imminent climate-driven coastal uninhabitability.

People vote with their feet—and their mortgages.

8. The Real Drivers of Settlement Patterns Are Economics and Infrastructure, Not Climate Fear

Let’s spell out the well-known, empirically supported factors that drive modern settlement shifts:

1. Suburban expansion

As cities globalize, growth typically moves outward, not inward.

2. Rising coastal land prices

As wealthy areas densify, lower-income residents move slightly inland.

3. New transportation corridors

Roads, railways, and airports often spur inland development.

4. Industrial relocation

Factories and logistics hubs often move inland to cheaper land.

5. Electrification and grid expansion

Satellite sensors detect light, not humans.

None of these are climate phenomena. All are consistent with ordinary economic development.

Until the authors can demonstrate that humans are abandoning coastal areas because of climate fears—and not simply developing new land inland—the claimed “climate retreat” remains an assumption.

Conclusion: The Study’s Narrative Outpaces Its Evidence

The press release frames global settlement shifts as a climate adaptation story, but the underlying evidence appears far too thin to support that narrative.

To summarize the core problems:

• Developers don’t base decisions on climate risk—they base them on profit.
• Nightlight data is an unreliable proxy for population movement, especially across countries with uneven electrification.
• Historical population data shows long-term growth on coastlines.
• Economic expansion inland is not the same as retreat from the coast.
• The study itself admits that vulnerability—not hazard increase—is the key driver.
• Wealthy nations continue to build aggressively on coastlines, contradicting claims of climate-driven retreat.

The real story appears to be one of development patterns, not climate exodus.

The press release presents an alarming climate-driven narrative, but the evidence it cites actually points to something else entirely:

People are not fleeing coasts—coasts are evolving, developing, and in many places, thriving.

And until studies like this validate their satellite interpretations with real-world population and development data, claims of global “retreat” should be viewed with healthy skepticism.