Forecasting the Next Flood Without Knowing Where the Rivers Go

Matt Wilson spent thirty years learning to read Western North Carolina's rivers. How the French Broad behaves during spring floods. Where the Swannanoa crests when mountain storms stall overhead. The precise relationship between rainfall in the headwaters and water levels downstream. A hydrologist with the Lower Mississippi Valley River Forecast Center, Wilson built models calibrated against decades of observation—the kind of patient, technical expertise that keeps mountain communities from drowning in their sleep.

Last September, Hurricane Helene erased that expertise in a single day.

The French Broad doesn't flow where it used to. Neither does the Swannanoa, the Pigeon, or dozens of smaller waterways threading through the mountains. Helene dropped over thirty inches of rain in some locations—what the NC State Climate Office called "off the charts"—and the water didn't just flood the rivers. It physically moved them. Streams that ran through narrow valleys now cut through what Wilson calls "small canyons." Some waterways changed their paths entirely, carving new channels through landscape that had held steady since the catastrophic floods of 1916.

The French Broad crested at 24.67 feet in Asheville on September 27, 2024, shattering the 108-year-old record by more than eighteen inches. Eighteen new flood records fell across North Carolina alone. The USGS estimated that dozens of their stream gauges observed peak heights that exceeded their measurement capacity. Some gauges were destroyed entirely.

Wilson is now trying to forecast floods on rivers he no longer understands, using models built for a physical system that doesn't exist anymore.

Years. Not months of recalibration. Years before he can forecast with the confidence his profession demands. The next storm season is already here.

Every time a weather system approaches the mountains, Wilson faces the same test. He looks at rainfall projections. He runs his models—models he knows are based on rivers that changed course. He issues flood warnings to communities in Asheville, Canton, and dozens of smaller towns along waterways that might not behave the way his thirty years of experience tells him they should. Then he waits to see if he was right.

They have to see the disaster before they can predict it. The next time heavy rain falls on the mountains, Wilson will be forecasting with models built on incomplete understanding of how these rivers work now. He knows this. The emergency managers who depend on his warnings know this. He issues the forecasts anyway, because someone has to.

His team is using Helene itself as "the new core set of data for Western North Carolina"—building models based on the storm that broke their old models. The physical changes are so extensive that even with Helene's data, they're still guessing. The rivers carved new channels. The streambeds changed elevation. The relationship between upstream rainfall and downstream flooding—the fundamental equation Wilson spent his career mastering—is different now in ways they're still trying to measure.

The USGS is still repairing damaged gauges and surveying high-water marks more than a year after the storm. They're trying to document what happened so they can understand what might happen next. Documentation takes time. The weather doesn't wait.

The next five years? More storms. More uncertainty. More moments when Wilson has to issue a forecast based on models everyone knows are provisional. The climate that shaped those old river channels isn't coming back. Whatever "new normal" emerges will itself keep shifting as warming continues. He's not recalibrating to a stable system. He's trying to model a moving target while communities depend on his warnings for survival.

Wilson doesn't talk about whether he's considered leaving the field or what he tells his family about the limits of his knowledge. His public statements are careful, technical, focused on the work of recalibration. But the situation he's describing is impossible: maintaining professional standards when the systems you've spent your career mastering change overnight. Protecting communities when your expertise has been rendered obsolete by forces beyond your control. Doing your job when you know your tools are broken.

The next major storm will test whether his rebuilt models work. Or it will reveal new gaps in understanding. Or it will change the rivers again, requiring another round of recalibration. Wilson won't know which until it happens.

He shows up. He runs the models. He issues the forecasts. He does the work despite knowing it might not be enough, because the alternative—telling mountain communities they're on their own—isn't an option he's willing to consider. The rivers changed course. The models broke. Wilson is still trying to figure out what the new normal is, one forecast at a time, hoping the next storm doesn't arrive before he does.