The evening of September 1, three days after Hurricane Ida made landfall in Louisiana, the remnants of the storm passed over New York City and dumped more than five inches of water on the metropolitan area in the span of a few hours. The resulting flood event stranded thousands of cars on the roadways, spewed water into countless basement apartments, and shut down almost the entire subway system. At least 13 people died in the city alone.

According to data from NOAA, the Ida event qualified as at least a “hundred-year flood,” which means it has around a 1 percent chance of happening in any given year. Except the event came only a few weeks after another once-in-a-lifetime flood; that storm broke the one-hour record for precipitation in New York at 1.94 inches, but the Ida event shattered that record again, bringing 3.14 inches of rain to Central Park in a single hour.

Descriptors like “hundred-year flood” aren’t just terms of art. They reflect how we understand and interpret the natural world, and when we witness events like Elsa and Ida, we can be confident that our understanding is sorely out of date. If we want to catch up to disasters like these, we have to come to terms with the fact that we need nothing short of a radical, full-scale transformation of the world we live in; to survive, we may need to alter it beyond recognition. We have to rethink how we build, live, and do almost everything, and rebuild most of the physical environment we take for granted.

Many of us are wont to think of climate change as the kind of thing that happens in particular, defined areas: Fires burn in California; hurricanes strike Florida; heat waves scald the deserts of Arizona. If it has done nothing else, the past summer has shown that this framework is false. The temperate Pacific Northwest burned beneath a heat dome, an apocalyptic flood raged through the hills of Middle Tennessee, and the Caldor Fire jumped the Sierra Nevada and burned through Lake Tahoe.

The increase in extreme rainfall events is of particular note, since it threatens many of the northern cities that we have thought of as refuges from climate change. Warmer air can hold more moisture, which means that storms passing through that air can gather and release more rain. This phenomenon is behind the convective weather patterns known as “rain bombs,” and although the science is not yet definite, it appears to affect larger storm systems like Ida and the rainstorm that struck Tennessee.

The sheer territorial extent of areas vulnerable to extreme weather has dire consequences: It’s now difficult to know where and how to concentrate our resources, and almost impossible for the eventual victims to prepare for impact. The Federal Emergency Management Agency publishes a nationwide database of flood zone maps that purport to show which areas are vulnerable to flooding. If you look at the maps for New York City, you will see that they bear almost no resemblance to the places where flooding took place during Ida: elevated parts of Brooklyn like Bushwick, Bed Stuy, and Park Slope were buried beneath several feet of water. The folly of the flood maps is the same as the folly of the hundred-year flood: They’re both attempts to delimit and define a crisis that seems to have no real outer bound.

Imagine you wanted to prevent the kind of flooding that happened in New York City during Ida. What would you do? First, you would have to find a way to soak up some of the rainwater as it falls on the city streets, because concrete and asphalt are very bad at absorbing water. One way to do this would be to create thousands or perhaps tens of thousands of natural water sinks such as bioswales or grasslands, but you’d have to make sure there were several in every single neighborhood, and you’d have to find a place to put them that didn’t interfere with private property or public rights-of-way. You could also revamp and expand the city’s storm drain system, or create a pump system to match the robust one in a city like New Orleans, but those interventions would cost untold billions of dollars, and there’s no guarantee they could keep up with the flooding during a rain event like Ida.

Let’s move on to residential damage. If you can’t fix the problem of drainage and water absorption, you need to get people out of low-lying homes, which means you have to find a way to relocate many of the 100,000-plus people who live in illegal basement and sub-grade apartments. This would entail addressing the endemic shortage of housing that forced many low-income residents to live in such converted units in the first place. There are also a million or more residents who live in or near the coastal floodplain, making them vulnerable to storm surge events like Sandy, and you’d have to find new homes for them as well. Or you could mandate that buildings in the floodplain be elevated above a certain height, but what about all the houses that are already there?

Then there’s the transportation system. The city is lucky to have few major expressways that run below street grade, but even the elevated expressways turn into swimming pools during Ida-caliber rain events, so you’ll have to modify them to create adequate drainage, then find a way to keep all that excess water away from nearby homes and streets. The subway, too, is an easy destination for all falling water, and in order to prevent major floods you will have to revamp the entire architecture of the underground train network. An estimated 20 percent of entrances are vulnerable to lethal flash flooding, so you’ll have to revamp those at a cost of who knows how many billion dollars. On a sunny day, the current system must pump around half a million gallons of water per hour in order to keep the stations dry; at its peak power, the Ida remnants dropped around 4 million gallons of water on Brooklyn in the space of a single hour, eclipsing the pumps’ capacity by around eightfold.

This thought exercise should make it clear: If you try to build your way out of a problem like climate change, you end up needing to throw out what you have and start again with something new. For a city with more than 8 million people and centuries of history, that is a difficult pill to swallow, and there are few if any politicians or public leaders who are interested in ingesting it. In places like Western Tennessee, where the population is less wealthy and infrastructure funding is harder to come by, this revolutionary rebuilding is even less practicable.

It is not easy to admit that such measures as these are necessary, and it is even harder to think about how we would go about accomplishing them. The alternative, though, is to live in a very dangerous world, a world that looks nothing like the one we thought we knew.