Why America's Biggest Cities Are All Sinking

Seven hundred kilometers above your head, a satellite passes over America again and again. Each time, it bounces radar off the ground and times the echo. Compare enough passes, and you can tell whether a patch of ground the size of a house lot has moved up or down by a few millimeters.

In 2025, a team led by Leonard Ohenhen at Columbia University's Lamont-Doherty Earth Observatory used that kind of satellite data to check every US city with more than 600,000 people. That's 28 cities. They expected to find a few sinking hot spots.

Instead, they found that all 28 are sinking. Not just New Orleans and Houston, which everyone expected, but Chicago, Denver, Columbus, Indianapolis, Detroit. Cities hundreds of miles from any ocean.

Guess first: Before you read on, which do you think is the biggest cause: the weight of buildings, rising seas, or something under the ground you can't see? Hold on to your answer.

Here's what the satellites saw:

• In every one of the 28 cities, at least a fifth of the land is sinking. In 25 of them, at least two-thirds is.
• About 34 million people live on that sinking ground, roughly one American in ten.
• Houston is the fastest. More than 40% of its area sinks faster than 5 millimeters a year, and 12% sinks twice that fast. Fort Worth and Dallas are close behind.
• About 29,000 buildings sit in zones where the ground is moving unevenly enough to threaten them.

That leaves three hard questions. Why would a city with no coast sink at all? Why is uneven sinking more dangerous than fast sinking? And which cities will stop? Let's work them out.

Hard question 1Why are cities far from any coast sinking too?

When we hear "sinking city," we picture the sea creeping in. But the sea has nothing to do with whether Denver's ground moves. Sinking is about what's happening under the city, and the study found three main culprits.

1. We drink the ground. Under most American cities lie aquifers: layers of sand, gravel and clay soaked with water. Cities and farms pump that water up. The researchers estimate that groundwater pumping accounts for about 80% of the sinking they measured. In Texas, pumping oil and gas from the ground adds to it.

Why does pumping water make the ground drop? Picture a water bed. The water holds you up. Drain some, and you sink. The water in the tiny spaces between grains of sand and clay is doing the same job: it carries part of the weight of everything above it. Take the water out, and the grains have to carry that weight alone. Sand barely notices. Clay squeezes like a sponge.

2. The ghost of the ice age. About 20,000 years ago, an ice sheet up to two miles thick covered much of northern North America. Its weight pressed the land beneath it down into Earth's slowly flowing interior, and that pushed the land around its edges up, the way a mattress bulges when you sit on the middle. Geologists call that rim a forebulge. The ice is long gone, so the bulge is slowly settling back down, by about 1 to 3 millimeters a year in some places. The study names New York, Indianapolis, Nashville, Philadelphia, Denver, Chicago and Portland among the cities affected.

3. The weight of the city itself. New York has more than a million buildings. A 2023 study suggested their combined weight helps press the ground down. It's a small effect compared to pumping, but it's real.

Now let's see what a "small" number does over a lifetime:

A foot is a lot. It's the difference between a storm drain that empties and one that backs up into the street.

The takeaway: inland cities sink for the same main reason coastal ones do: we pump water out from under them. The ice age adds a slow settling in some places, and the sea is only what makes sinking dangerous on the coast. It isn't what causes it.

Hard question 2Why is uneven sinking worse than fast sinking?

Guess first: If your whole house sank one foot, evenly, would it crack?

Surprisingly, probably not. If every part of a building moves down together, the building just rides along. The walls don't bend. The doors still close. You might not notice at all, except that your driveway now slopes a little differently.

The trouble starts when one end sinks faster than the other. Now the building has to bend, and brick, concrete and pipes don't like bending. Engineers measure this as angular distortion: how far one point drops compared to another, divided by the distance between them. As a common rule of thumb, around 1 in 300 is where cracks often start showing in walls; 1 in 500 is a typical safe limit.

Let's test a building 30 meters long (about 100 feet). Suppose one end sinks 5 mm a year and the other only 2 mm a year:

That's right at the edge where cracks appear, even though neither end is sinking especially fast. Now compare a building where both ends sink 10 mm a year: twice as fast, and zero distortion.

This is why the study's most important number isn't the speed. It's the roughly 1% of city land where neighboring patches move at different rates, holding about 29,000 buildings. By share of buildings in those zones, San Antonio (about 1 in 45), Austin (1 in 71), Fort Worth (1 in 143) and Memphis (1 in 167) came out on top. New York has hot spots too, around LaGuardia Airport, Rikers Island, parts of Jamaica Bay and Staten Island.

The takeaway: a building can ride out a lot of even sinking. What breaks things is the difference in sinking from one spot to the next, so the slow-but-uneven places can be riskier than the fast-but-even ones.

Hard question 3Which cities will stop sinking?

Guess first: If most sinking comes from pumping groundwater, what happens when a city stops pumping?

Here's the hopeful part. The satellites found three cities, Jacksonville, Memphis and San Jose, where areas of rising ground more than make up for the sinking. The researchers suggest the rise may come from aquifers refilling near rivers and other water sources.

San Jose is a famous example. In the mid-1900s, heavy pumping in the Santa Clara Valley dropped parts of downtown San Jose by many feet. The valley then brought in surface water and refilled its aquifers on purpose, and the sinking largely stopped by the 1970s. Houston did something similar after 1975, and its old downtown core slowed dramatically (more on that in our Houston article).

But there's a catch, and it's the key to the whole story. When clay is squeezed, much of the squeeze is permanent. Clay grains get pressed into a tighter arrangement, and refilling the water doesn't push them back apart. So the prediction is:

• Cities that cut pumping and recharge their aquifers can stop sinking, and sandy areas may even rise a little.
• The ground they already lost stays lost. A city can stop the clock, but not rewind it.
• The ice-age settling won't stop on any human timescale. It's a few millimeters a year, and it will go on for thousands of years.

The takeaway: sinking from pumping can be slowed or stopped, and that's already happened in places. But what's lost mostly stays lost, which is why measuring early matters so much.

Remember your guess from the top? If you said "something under the ground you can't see," you were right: the water.

More questions people ask

Try it yourself

Fill a sponge with water, set it on a plate, and place a book on top. Measure the stack. Now press gently on the sponge's side to let water out, and measure again. The stack dropped. Fill the sponge back up: it won't quite return to where it was if you squeezed it hard. That's clay under a city, in miniature.

Watch next

Why Houston Is America's Fastest-Sinking City

Sources