Beneath the streets of hundreds of cities, the earth is quietly giving way — a slow, largely invisible crisis that satellite science has now brought into focus for the first time. Over nine years, researchers tracked land subsidence across 465 cities in 107 countries, finding that some urban neighborhoods are sinking fast enough to fracture pipes, warp buildings, and deepen flood vulnerability with each passing year. The study, published in Nature Cities, does not merely document a problem — it offers, for the first time, a common language and a shared map with which the world's cities might b
Satellite data reveal widespread land subsidence threatening major cities globally
Land subsidence damages buildings, foundations, and infrastructure systems, increasing flood risks for urban populations in affected cities.
The ground beneath major cities is sinking, and no one had a clear picture of where.
Mark
So we have satellite data showing which cities are sinking fastest. But what actually causes the ground to drop in one place and stay stable in another?
Mimi
It's usually groundwater extraction—cities pump water out of aquifers faster than it recharges, and the soil compacts. In some places it's the weight of development itself, or subsurface mining. The causes vary, but the effect is the same: the ground gives way.
Luke
The study doesn't break down causes by city, though, right? We know Medan is sinking 33 millimeters a year, but we don't know from this data alone whether that's groundwater depletion, or something else entirely.
Mimi
Correct. The satellite data shows the movement, not the mechanism. That's where local investigation comes in. But having the map is the first step—you can't fix what you don't measure.
Mark
And the cities that are stable—New York, London, Tokyo—what does that tell us?
Luke
It tells us subsidence isn't inevitable in dense urban areas. These cities have either managed their groundwater better, or they're built on geology that doesn't compact easily, or both. But we'd need to know more about each one to say why.
Mimi
It's also worth noting that nine years is relatively short for detecting long-term trends. A city that looks stable now might start sinking if extraction patterns change.
Mark
So this is really a baseline. A starting point for watching.
Mimi
Exactly. The value isn't just in knowing which cities are sinking fastest today. It's in having a consistent method to track whether things get better or worse.
Il Polso
- Cities from Mexico City to Medan are losing ground at rates that physically deform infrastructure — in the worst cases, nearly nine inches of sinking in a single year.
- For decades, fragmented methods and isolated studies left policymakers without a reliable way to compare risk across cities, making coordinated response nearly impossible.
- A nine-year satellite analysis has now produced the first systematic global benchmark, giving urban planners a consistent, cross-comparable picture of where the ground is moving and how fast.
- Some of the world's most prominent cities — New York, Tokyo, London — show near-stable ground, proving that severe subsidence is not inevitable but concentrated and often human-driven.
- Researchers caution that administrative boundaries, sparse monitoring networks, and the limits of a nine-year window mean the full picture remains incomplete — and some stable trends could yet accelerate.
Beneath the streets of hundreds of cities, the earth is quietly giving way — a slow, largely invisible crisis that satellite science has now brought into focus for the first time. Over nine years, researchers tracked land subsidence across 465 cities in 107 countries, finding that some urban neighborhoods are sinking fast enough to fracture pipes, warp buildings, and deepen flood vulnerability with each passing year. The study, published in Nature Cities, does not merely document a problem — it offers, for the first time, a common language and a shared map with which the world's cities might begin to respond.
The ground beneath hundreds of the world's cities is sinking — and until a sweeping new satellite study, no one had a reliable map of where, or how fast. Researchers processed nine years of high-resolution imagery spanning 465 cities across 107 countries, producing what Nature Cities published as the first truly systematic global assessment of urban land subsidence. The stakes are not abstract: ground that drops several inches a year cracks building foundations, ruptures sewer lines, and steadily raises the flood risk for the people living above it.
The crisis is unevenly distributed. Mexico City's most vulnerable neighborhoods sank roughly 217 millimeters in a single year — about eight and a half inches — though the city's median rate was a far slower 8 millimeters annually. The most severe median sinking rates clustered in Asia: Medan, Indonesia led at 33 millimeters per year, followed by Bandung at 22 millimeters, and Isfahan, Iran at 11 millimeters. Meanwhile, cities like New York, Tokyo, London, and Beijing showed movement of less than 2 millimeters annually — a reminder that subsidence is not fate, but is often driven by specific, addressable causes such as groundwater extraction or the compressive weight of dense development.
The researchers are measured about what the data can and cannot prove. Subsidence just beyond city administrative boundaries may go uncounted. Monitoring gaps widen uncertainty in data-sparse regions. And nine years, while meaningful, remains a snapshot — stable trends might shift, alarming ones might ease. What the study does provide is a foundation: cities now know where to look, where to measure more closely, and where intervention might prevent collapse. The work ahead is sustained watching — and building urban systems resilient enough to adapt to ground that can no longer be taken for granted.
The ground beneath major cities around the world is sinking, and until now, no one had a clear picture of where or how fast. Researchers analyzing nine years of satellite data have mapped land subsidence across 465 cities in 107 countries, revealing a problem that threatens the foundations of urban life—literally. The work, published in Nature Cities, shows that some neighborhoods are dropping several inches per year, a pace that warps buildings, fractures sewage lines, and pushes flood risk higher with each passing season.
The challenge has always been fragmentation. Previous studies of ground subsidence used different measurement methods, covered different time periods, or examined only a handful of cities at a time. Policymakers trying to decide where to invest in infrastructure reinforcement or drainage upgrades had no reliable way to compare one city's risk to another's. Shengli Tao, Xie Hu, and their colleagues set out to change that by processing high-resolution satellite imagery collected between 2014 and 2023, creating the first systematic global assessment of how and where urban ground is moving.
The data tells a story of uneven crisis. In Mexico City, the fastest-sinking areas dropped roughly 217 millimeters—about 8.5 inches—in a single year, though the citywide median rate was far slower at 8 millimeters annually. Torreón, also in Mexico, sank at 17 millimeters per year. These are not abstract numbers. A building sitting on ground that drops 8 inches a year will crack. Pipes will rupture. The geometry of the city changes. But the most severe subsidence clustered in Asia. Medan, Indonesia experienced the highest median sinking rate at 33 millimeters per year. Bandung, also in Indonesia, dropped 22 millimeters annually. Isfahan, Iran, sank at 11 millimeters per year. These cities are losing ground faster than most places on Earth.
Nine major cities—Boston, New York, London, Beijing, Tokyo, Seoul, Sydney, Montevideo, and Naples—showed relatively stable ground, with movement ranging from barely measurable to less than 2 millimeters per year. The stability in these places offers a baseline: subsidence is not inevitable. It is concentrated, localized, and often driven by specific causes—groundwater extraction, soil compaction, or the weight of development itself.
The researchers are careful about what their data can and cannot show. The study measured relative movement within large city boundaries, which means it may miss subsidence happening just beyond administrative lines. In regions where ground-monitoring stations are sparse, the uncertainty widens. A nine-year window, while substantial, is still a snapshot. Trends that appear stable might accelerate. Patterns that seem dire might reverse. What the data does provide is a foundation for action. Cities now know where to look, where to measure more closely, and where urgent intervention might prevent infrastructure collapse. The next step is sustained monitoring—watching whether these sinking rates hold steady, accelerate, or slow, and using that knowledge to build cities that can adapt to ground that is no longer solid.
Citazioni salienti
Previous assessments used different methods or time periods, making global comparisons difficult; this study provides the first systematic cross-city analysis.— Research team, Nature Cities
Vuoi la storia completa?
Leggi l'originale su Phys.org ↗
Science & Space